Vol. 63, Issue 1, 147-158, January 2003

Cell-Level Pharmacokinetic Model of Granulocyte Colony-Stimulating Factor: Implications for Ligand Lifetime and Potency in Vivo

Casim A. Sarkar¹ and Douglas A. Lauffenburger

Department of Chemical Engineering (C.A.S., D.A.L.), Biotechnology Process Engineering Center (C.A.S., D.A.L.), Department of Biology (D.A.L.), and Biological Engineering Division (D.A.L.), Massachusetts Institute of Technology, Cambridge, Massachusetts

The cytokine granulocyte colony-stimulating factor (GCSF) is of great clinical importance, with primary application to rapidly elevate the peripheral neutrophil levels of chemotherapy patients through accelerated granulopoiesis. However, these mature bloodstream neutrophils express the GCSF receptor (GCSFR), presenting a significant and specific clearance mechanism of circulating GCSF that increases with time. Here, we formulate a mathematical model that describes these cell-level GCSF/GCSFR dynamics and correlate the effect of these endocytic trafficking processes to ligand depletion in an in vitro culture. We further incorporate this cell-level model into an existing pharmacokinetic/pharmacodynamic (PK/PD) model, to gain insight into the effects that specific molecular and cellular parameters may have on overall PK/PD effects in vivo. Our cell-level model suggests that ligand depletion may be reduced in vitro by decreasing the endosomal affinity of endocytosed GCSF/GCSFR complexes, matching experimental findings. Additionally, our modified PK/PD model suggests that a GCSF analog with a modification that effectively eliminates renal clearance should have a significantly longer half-life in vivo and should therefore improve peripheral neutrophil counts. This is consistent with clinical studies on a polyethylene glycol chemical conjugate of GCSF termed SD/01. The model predicts that a GCSF analog that eliminates renal clearance and has reduced endosomal binding affinity may result in an even longer ligand half-life and increased neutrophil counts at a lower dose than either wild-type GCSF or SD/01. More generally, this type of hierarchical model provides a correlation between the molecular and pharmacological properties of a drug and may elucidate design goals for such protein therapeutics.