Abstract

The clastogenic potential of a recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) therapeutic protein formulation was assessed in human peripheral blood lymphocyte (HPBL) cultures. Final formulation, rather than pure protein, was evaluated to simulate exposure conditions encountered in man. Because the formulation excipients (citric acid, sodium phosphate, mannitol, human serum albumin and polyethylene glycol) were found to alter cell-cycle kinetics and interfere with S-9 metabolic activation in vitro, a novel testing sequence and protocol were used to distinguish between rhGM-CSF and excipient effects and to guard against false negative results. Initial trials were performed to establish the volume of formulation alone (with and without rhGM-CSF) that would not cause severe cell-cycle delay, interfere with S-9 metabolic activation or inhibit positive control responses. This maximum tolerated volume of formulation was then incorporated in all dosed and control groups of each respective assay phase to give the same extent of cell-cycle delay. In the main assay, 24-hr HPBL cultures were exposed to dose levels of 0, 50, 100, 200 and 350 mug rhGM-CSF/ml for 24 hr in -S-9 phases and 0, 200, 350, 500 and 650 mug/ml for 2 hr in +S-9 phases, and harvested 24 hr later. No significant increases in the percentage of cells with structural chromosomal aberrations were found in either test phase. These results show that rhGM-CSF is not clastogenic in HPBL at greater than 17,000 times human exposure levels, and demonstrate that valid cytogenetic assay data with formulations containing cytotoxic excipient can be obtained in vitro.