Modelling and sensitivity analysis of urinary platinum excretion in anticancer chemotherapy for the recovery of platinum

Abstract

Platinum (Pt) based antineoplastics are important in cancer therapy. To date the Pt which is urinary excreted by the patients ends up in wastewater. This is disadvantageous from both an economic as from an ecological point of view because Pt is a valuable material and the excretion products are toxic for aquatic organisms. Therefore, efforts should be made to recover the Pt. The urinary excretion of Pt from two antineoplastics are taken under consideration, i.e. cisplatin and carboplatin. Using these reference compounds, a scenario analysis based on administration statistics from Ghent University Hospital in combination with compartmental models for urinary Pt excretion was performed to simulate the average Pt excretion profile during common treatment schemes. These average profiles can be used to assess the technical, social and economic feasibility of Pt-recovery from urine or wastewater. A one-compartment model is used for cisplatin, which is calibrated using the experimental data of six patients. In contrast, a two-compartment model with parameters from literature is used for carboplatin. A Global Sensitivity Analysis revealed kel, the rate constant of elimination, is the most sensitive parameter in the one-compartment model whereas Qu, the urine production rate, was the most sensitive in the two-compartment model for the Pt concentration Cu in urine and the excreted mass of Pt via urine. A GLUE uncertainty analysis showed that all experimental data are within the 95% uncertainty boundaries.