Japanese Journal of Clinical Oncology 31:179-184 (2001)
© 2001 Foundation for Promotion of Cancer Research
Population Pharmacokinetic Analysis of Cisplatin and Its Metabolites in Cancer Patients: Possible Misinterpretation of Covariates for Pharmacokinetic Parameters Calculated from the Concentrations of Unchanged Cisplatin, Ultrafiltered Platinum and Total Platinum
1Department of Biopharmaceutics, Meiji Pharmaceutical University, Tokyo and 2Department of Internal Medicine, National Kinki Central Hospital for Chest Diseases, Osaka, Japan
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Background: Usually, total and filtered platinum concentrations in plasma are monitored after cisplatin administration. However, these concentrations represent a mixture of unchanged cisplatin and metabolites. In this work, we studied population pharmacokinetic analysis based on these platinum concentrations.
Methods: Twenty-seven patients (23 males, four females) were administered cisplatin (60–100 mg/m2) with intravenous constant infusion for 90 min. Blood samples were taken at about three points per patient. The concentrations of cisplatin and platinum in the plasma were determined by high-performance liquid chromatography and atomic absorption spectrometry, respectively. Population pharmacokinetic analysis was performed using the program NONMEM (Version V) with the one- or two-compartment model with zero-order infusion.
Results: The clearance and volume of distribution for all platinum species studied were significantly related to the body surface area of the patients. Only the clearance of filtered platinum was significantly related to urinary N-acetyl-ß-D-glucosaminidase and the other covariates were not related to these pharmacokinetic parameters with respect to unchanged cisplatin and total platinum concentrations.
Conclusion: The dosage regimen based on the filtered platinum concentration which is usually monitored may result in possible misinterpretation because the detected covariate is different between unchanged cisplatin and filtered platinum.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Cisplatin (CDDP) is a potent anticancer drug used in the treatment of various types of cancer. However, severe nephrotoxicity limits the dose and frequency of CDDP therapy (1).
CDDP undergoes ligand-exchange reactions, which are virtually irreversible (2,3). In biological fluids, CDDP is transformed immediately into aquated CDDP as a result of the release of chloride ion and equilibrium between CDDP and its aquated form is maintained. Aquated and unchanged CDDP also react readily with nucleophiles (3–5). CDDP is biotransformed through binding to low molecular mass substances (such as glutathione, methionine and cysteine) and high molecular mass substances (such as albumin and nucleotides) and the resulting metabolites are called ‘mobile’ and ‘fixed’ metabolites, respectively (5). The pharmacokinetic characteristics of these platinum species are different because their structures are different (6,7). The CDDP in plasma is instantly eliminated and mobile and fixed metabolites are gradually increased. The mobile metabolites in plasma are eliminated more slowly than that of CDDP and fixed metabolites are little decreased after CDDP and mobile metabolites have been eliminated (6,7).
In order to monitor the efficacy and safety of drug treatment, it is important to determine the concentration(s) of the active species. When CDDP and sodium thiosulfate are administered concomitantly, the nephrotoxicity estimated by blood urea nitrogen (BUN) is ameliorated. When CDDP is administered with sodium thiosulfate, the plasma filtered platinum (mixture of CDDP and mobile metabolites) concentration changes little, but the plasma CDDP concentration decreases dramatically, because CDDP binds strongly to sodium thiosulfate (8). This result suggests that the effect of sodium thiosulfate is associated with the alteration of the pharmacokinetics and pharmacodynamics of CDDP. Furthermore, we also showed that mobile metabolites in plasma were less toxic than CDDP after administration of mobile metabolites to rats and suggested that intact CDDP in plasma is the major species responsible for nephrotoxicity after CDDP administration (7). Therefore, it is important to separate the platinum species into CDDP and mobile metabolites to monitor their pharmacokinetics and pharmacodynamics after CDDP therapy.
In the last decade, the population approach has been widely used in clinical pharmacokinetics as an effective technique, because it can be used to obtain population mean values, including regression relationships with specific pathophysiological characteristics, together with estimates of interindividual and residual variability (9–11). Recently, Nagai et al. reported the population pharmacokinetics and pharmacodynamics of CDDP in patients with cancer (11). They showed that the clearance of CDDP is changed with the body surface area (BSA) of patients and infusion rate of CDDP. As mentioned above, however, many investigators have studied the pharmacokinetics of CDDP by measuring the concentrations of mixed platinum species, such as total (including all platinum species) and filtered platinum (including CDDP and its mobile metabolites). If the covariates of the pharmacokinetics and pharmacodynamics of CDDP are estimated by the concentrations of mixed platinum species, interpretation of these concentration–time profiles should be carefully considered. Therefore, it is important to evaluate exactly the apparent covariate factors when these pharmacokinetic factors are estimated after CDDP administration. In the present study, we studied the population approach for the concentrations of CDDP, ultrafiltered platinum and total platinum in patients with cancer.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Chemicals
CDDP for preparing quantitative standard solutions was kindly donated by Nippon Kayaku (Tokyo, Japan). Other chemicals were used of analytical grade.
Patients and Data Description
A total of 27 patients (23 male, four female) in National Kinki Central Hospital received an intravenous infusion of CDDP. This study was approved by the Hospital Ethics Committee Board at National Kinki Central Hospital and all patients gave informed consent before the study. Table 1 gives a summary of the patients’ data.
|
All patients were treated with combination chemotherapy. They received etoposide, vinca alkaloid, irinotecan or mitomycin C with CDDP. Body surface area (BSA), gender, age, GOT, GPT,
-glutamyl transpeptidase (GTP), alkaline phosphatase (ALP), lactate dehydrogenase (LDH), blood urea nitrogen (BUN), serum creatinine (SCR), white blood cell count (WBC), albumin (ALB), 
1-acid glycoprotein (AGP), urinary ß2-microglobulin (UBMG) and urinary N-acetyl-ß-D-glucosaminidase (UNAG) were monitored.
Assay
The concentrations of CDDP in plasma were determined according to a high-performance liquid chromatographic (HPLC) method with on-line post-column derivatization, as reported previously (6,12). The concentrations of platinum in plasma were determined by atomic absorption spectrometry.
Population Pharmacokinetic Analysis
The population pharmacokinetic parameters were calculated on a Dell Dimension XPS R450 computer (Pentium II, 450 MHz, 128 MB RAM) with the NONMEM program (Version V) (13). As the first step in the model building, the basic structure model was selected. A one-compartment open model with zero-order administration and first-order elimination (PREDPP program, subroutine ADVAN1 and TRANS2) was used to describe the plasma concentration–time course of CDDP and ultrafiltered platinum. A two-compartment open model with zero-order administration (PREDPP program, subroutine ADVAN3 and TRANS2) was used for total platinum.
The statistical models with proportional error were selected by the examination of the objective function (OBJ) according to the following equations:
CLj =
CLj)
Vdj =
Vdj)
Qj =
Qj)
Vssj =
Vssj)
Cij =
ij)
where CLj, Vdj, Qj, Vssj and CL, Vd, Q and Vss are the estimated values in the jth individual and the population means for total clearance (CL), volume of distribution (Vd), intercompartmental clearance (Q) and volume of distribution at steady state (Vss), respectively, and CLj, Vdj, Qj and Vssj are the interindividual variability in CL, Vd, Q and Vss, respectively. Cij and
ij is the residual variability in plasma platinum concentration. In the screening for significant fixed effect parameters, body surface area (BSA) was first evaluated as body size with the following equations:
CLj = 
1·BSAj3
and
Vdj = 2·BSAj3
where BSAj is BSA of the jth individual. Other patient characteristics (factors) such as SCR and BUN were evaluated with the following equations:
P = 1·BSAj·(1 + 2·Factor)
or
P = 1·BSAj·(1 + 2/Factor)
where P represents pharmacokinetic parameters such as CL and Vd. The changes in the OBJ values before and after adding one factor to the model were compared. When the factor showed more than a 3.84-fold improvement in the OBJ values (P < 0.05, chi-squared with one degree of freedom), the factor was significantly related to the kinetic parameter as a fixed effect in the model.
The scatter plots of the relationships between the observed and the predicted plasma platinum concentrations were presented visually using the final estimated population pharmacokinetic parameters.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Fig. 1 shows the time courses of plasma concentrations of CDDP, filtered and total platinum after 1.5 h of constant infusion of CDDP. The time courses of CDDP and filtered platinum showed similar patterns but that of filtered platinum was relatively variable compared with that of CDDP. On the other hand, elimination of total platinum showed two different phases and was very slow. This slow elimination rate is due to the fixed metabolites, which bound CDDP to high molecular mass biological substances.
|
The proportional error as the basic statistical model for analyzing the concentrations of CDDP, filtered and total platinum was selected because a better fit was observed than with the additive error as estimated by OBJ values (data not shown).
First, CL and Vd were modeled as a function of BSA because the drug dosage for cancer chemotherapy is generally used as a function of BSA. The CL and Vd for all platinum species studied were significantly related to BSA (2 log likelihood difference (l.l.d.) = 8.65; p < 0.005).
Table 2 shows the hypothesis tested concerning the intersubject variability of CDDP clearance. The OBJ value after adding BSA to the basic model was significantly improved (l.l.d. = 18.06), as has been reported previously. The CL and Vd in this model were 15.6 l/h/m2 and 12.2 l/m2, respectively, with interindividual variabilities of 3.36% (CL) and 36.5% (Vd) and a residual variability of 16.9%. The observed and predicted plasma CDDP concentrations were well correlated (Fig. 2A). However, none of the patients’ characteristics other than BSA gave a significant improvement in the model fitting. Nagai et al. (11) reported that in the final pharmacokinetic model of CDDP, only BSA and dose schedule affected the CL of CDDP. In our study, the dose schedule of CDDP did not alter, so other patients’ characteristics may not be detected. The elimination of CDDP via the kidneys is about 20–30% of the dose in patients and the other elimination pathway of CDDP is the biotransformation to mobile and fixed metabolites, which are platinum species bound to low and high molecular mass biological substances, respectively. Therefore, the predictable covariates that affected the total clearance of CDDP are renal function and biological substance concentrations. CDDP undergoes glomerular filtration, active secretion and probably reabsorption in patients (14). However, the renal elimination of CDDP is relatively low and the patients who participated in this study had relatively normal renal function and, therefore, the effects of renal function on the total clearance of CDDP could not be detected.
|
|
Table 2 shows the hypothesis tested concerning intersubject variability of filtered platinum clearance. The OBJ value after adding BSA to the basic model was significantly improved (l.l.d. = 10.68). In the pharmacokinetic analysis for filtered platinum, only urinary N-acetyl-ß-D-glucosaminidase (UNAG) gave a significant improvement compared with the basic model with BSA added (l.l.d. = 14.29). The filtered platinum contains unchanged CDDP and mobile metabolites, which bound CDDP to low molecular mass biological substances. As mentioned above, CDDP is excreted via the kidneys at about 20–30% of the dose in patients, whereas mobile metabolites are excreted about at 50% in rats (7). The mobile metabolites undergo glomerular filtration, active secretion and probably active reabsorption in the kidney (15). Urinary NAG is thought to be a marker of disturbance of the renal proximal tubule, suggesting that renal clearance of mobile metabolites may be affected by the function of the renal proximal tubule represented by UNAG. CL and Vd in the final model were 13.6 x (1–0.46/UNAG) l/h/m2 and 10.3 l/m2, respectively, with interindividual variabilities of 14.4% (CL) and 55.8% (Vd) and a residual variability of 20.7%. These variabilities were relatively higher than those of CDDP. The UNAG level studied ranged from 1.2 to 62.8 U/day and the calculated CL ranged from 13.6 x 0.62 to 13.6 l/h/m2. Although no patients with relative low renal function participated in this study, the CL may conversely be decreased if the renal function was low. Therefore, in the final model the filtered platinum may not have a clinically significant meaning. These results indicate that in the analysis of filtered platinum, the determining factor of clearance is represented by that of mobile metabolites. On the other hand, the determining factor of this low clearance in the analysis of total platinum concentration is represented by that of fixed metabolites because the elimination of fixed metabolites was fairly slow and CDDP and mobile metabolites were not detected in the elimination phase.
Table 3 shows the hypothesis tested concerning intersubject variability of total platinum clearance. The OBJ value after adding BSA to the basic model was significantly improved (l.l.d. = 11.42). However, no other patients’ characteristics gave a significant improvement in the model fitting as well as that with CDDP.
|
In conclusion, we showed that the dosage regimen based on the usual monitoring of the filtered platinum concentration may result in possible misinterpretation because the detected covariates which affected clearance differ between unchanged CDDP and filtered platinum. Therefore, we recommend that HPLC analysis is essential for the evaluation of the pharmacokinetics and pharmacodynamics of cisplatin.
|
FOOTNOTES |
|---|
+ For reprints and all correspondence: Hiroyasu Ogata, Department of Biopharmaceutics, Meiji Pharmaceutical University, 2–522–1 Noshio 2-chome, Kiyose, Tokyo 204-8588, Japan. E-mail: hiroogat@my-pharm.ac.jp
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
1 Deconti RC, Toftness BR, Lange RC, Creasey WA. Clinical and pharmacological studies with cis-diamminedichloroplatinum(II). Cancer Res 1973;33:1310–5.
2 Daley-Yates PT, McBrien DCH. Cisplatin metabolites in plasma, a study of their pharmacokinetics and importance in the nephrotoxic and antitumour activity of cisplatin. Biochem Pharmacol 1984;33:3063–70.[Web of Science][Medline]
3 Nagai N, Okuda R, Kinoshita M, Ogata H. Decomposition kinetics of cisplatin in human biological fluids. J Pharm Pharmacol 1996;48:918–24.[Web of Science][Medline]
4 Dedon PC, Borch RF. Characterization of the reactions of platinum antitumor agents with biologic and nonbiologic sulfur-containing nucleophiles. Biochem Pharmacol 1987;36:1955–64.[Web of Science][Medline]
5 Farris FF, King FG, Dedrick RL, Litterst CL. Physiological model for the pharmacokinetics of cis-dichlorodiammineplatinum(II) (DDP) in the tumored rat. J Pharmacokinet Biopharm 1985;13:13–40.[Web of Science][Medline]
6 Kinoshita M, Yoshimura N, Ogata H, Tsujino D, Takahashi T, Takahashi S, et al. High-performance liquid chromatographic analysis of unchanged cis-diamminedichloroplatinum (cisplatin) in plasma and urine with post-column derivatization. J Chromatogr 1990;529:462–7.[Web of Science][Medline]
7 Hanada K, Ninomiya K, Ogata H. Pharmacokinetics and toxicodynamics of cisplatin and its metabolites in rats: relationship between renal handling and nephrotoxicity of cisplatin. J Pharm Pharmacol 2000;52:1345–53.[Web of Science][Medline]
8 Nagai N, Hotta K, Yamamura M, Ogata H. Effects of sodium thiosulfate on the pharmacokinetics of unchanged cisplatin in rats and on the distribution of platinum species in rat kidney: protective mechanism against cisplatin nephrotoxicity. Cancer Chemother Pharmacol 1995;36:404–10.[Web of Science][Medline]
9 Sheiner LB, Beal SL. Evaluation of methods for estimating population pharmacokinetic parameters. III. Monoexponential model: routine clinical pharmacokinetic data. J Pharmacokinet Biopharmacol 1983;11:303–19.[Web of Science][Medline]
10 Sheiner LB. The population approach to pharmacokinetic data analysis: rationale and standard data analysis methods. Drug Metab Rev 1984;15:153–71.[Web of Science][Medline]
11 Nagai N, Ogata H, Wada Y, Tsujino D, Someya K, Ohno T, et al. Population pharmacokinetics and pharmacodynamics of cisplatin in patients with cancer: analysis with the NONMEM program. J Clin Pharmacol 1998;38:1025–34.
12 Hanada K, Nagai N, Ogata H. Quantitative determination of unchanged cisplatin in rat kidney and liver by high-performance liquid chromatography. J Chromatogr B 1995;663:181–6.[Web of Science][Medline]
13 Beal SL, Sheiner LB. NONMEM User’s Guide. San Francisco: NONMEM Project Group, University of California at San Francisco 1992.
14 Daley-Yates PT, McBrien DC. Cisplatin metabolites: a method for their separation and for measurement of their renal clearance in vivo. Biochem Pharmacol 1984;32:181–4.
15 Reece PA, Stafford I, Davy M, Morris R, Freeman S. Influence of infusion time on unchanged cisplatin disposition in patients with ovarian cancer. Cancer Chemother Pharmacol 1989;24:256–60.[Web of Science][Medline]
Received November 16, 2000; accepted January 25, 2001.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Minami, Y. Ohe, S. Niho, K. Goto, H. Ohmatsu, K. Kubota, R. Kakinuma, Y. Nishiwaki, H. Nokihara, I. Sekine, et al. Comparison of Pharmacokinetics and Pharmacodynamics of Docetaxel and Cisplatin in Elderly and Non-Elderly Patients: Why Is Toxicity Increased in Elderly Patients? J. Clin. Oncol., July 15, 2004; 22(14): 2901 - 2908. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||