A computational model of human granulopoiesis to simulate the hematotoxic effects of multicycle polychemotherapy.

Moderate intensification of conventional multicycle chemotherapy has recently been shown to improve treatment results in malignant lymphomas and might prove to be beneficial also in other malignancies. However, the feasibility of such regimens is mainly limited by their granulopoietic toxicity. To identify and quantify the basic cell kinetic mechanisms of damage and stimulation caused by cytotoxic drugs and recombinant human granulocyte colony-stimulating factor (rhG-CSF), respectively, we developed a mathematical model of human granulopoiesis that allows simulation of leukocyte concentration profiles under 10 different multicycle polychemotherapy regimens with varying drug composition, dosage, and scheduling, including rhG-CSF assistance. Clinical data on leukocyte profiles were obtained from large numbers of patients treated within several multicenter trials. Simulation studies show that the leukocyte profiles of all regimens can be appropriately fitted using one single set of assumptions and parameters for the cell kinetic effects of cytotoxic drugs and rhG-CSF. Furthermore, the model can be used to explain the interindividual heterogeneity of hematotoxicity by a differential chemosensitivity, which might be useful in drug scheduling for specific risk groups. It is demonstrated that the model can be used to design and to select new drug schedules for subsequent clinical trial testing.