Abstract
PEG-Asparaginase is an important part of many treatment protocols for ALL. In many centres Asparaginase activity is measured after administration of PEG-asparaginase. However, a predictive pharmacokinetic model is lacking. Such a model would be helpful for dose adjustment and decision making when to switch to another preparation due to the development of inactivating antibodies. Previously described models like linear one-compartment [V.I. Avramis et al., Blood 2002, 99: 1986–1994] or a one-compartment Michaelis-Menten model [H.J. Mueller et al., Cancer Chemother. Pharmacol. 2002, 49, 149–154] describe the data sufficiently for one dose alone, but cannot account for the phenomenom that the time to reach a lower activity limit after administration is not increasing with increasing the dose. Therefore, we analysed 1189 serum activity measurements in 185 children from the ALL-BFM 95 study. Patients received 500, 1000 or 2500 U/m2 PEG-Asp on up to 9 occasions. Serum asparaginase activity was measured using a semi-automatic enzymatic assay with a limit of quantification of 2 U/l [C. Lanvers et al. Anal. Biochem. 2002, 309, 117–126]. Data analysis was done using nonlinear mixed effects modelling (NONMEM Vers. V). Different models like Michaelis-Menten, linear first-order, Weibull and gamma models were tested. The best model applicable to all dosing groups was a modified first-order one-compartmental model with clearance increasing with time according to the formula: Cl = Cli*exp(0.0853*t) with Cli=initial clearance, and t=time. Addition of a second compartment did not improve the model. A typical activity-time course of a patient receiving 1000 U/m2 is shown below displaying the typical shape observable in all patients and in all doing groups. The population parameters found were: Volume of distribution (V) 1.05 ± 27.3% l/m2, Cli 60.3 ± 70.8% ml/day/m2 (mean ± interindividual variability). Interoccasion variability was substantial with 0.223 l/m2 for V and 37.7 ml/day/m2 for Cl, respectively. A subgroup of one third of the patients is identifiable showing a high clearance probably due to the development of inactivating antibodies. Drug monitoring of serum PEG-Asparaginase activity is required to identify these patients who do not benefit from PEG-Asp therapy. The pharmacokinetic model presented here should help to reduce the number of required serum samples per patient.
Figure Figure
Macroscopic and Multi-Scale Models for Multi-Class Vehicular Dynamics with Uneven Space Occupancy: A Case Study