Myeloablation with diaziquone: in vitro assessment.

UNLABELLED The promising antineoplastic agent diaziquone is associated with prolonged aplasia and rare instances of bone marrow necrosis, but only mild extramedullary toxicity. To explore the drug's potential as a myeloablative agent prior to bone marrow transplantation, we compared its effects on hematopoietic versus marrow stromal cells. After short-term (one to six hours) or prolonged (three to seven days) exposure to the drug, marrow was assayed for hematopoietic (CFU-Mix, BFU-E, CFU-GM) and stromal (CFU-F) colony-forming cells and studied in long-term marrow culture (LTMC). One- and three-hour treatments produced little cytotoxicity, even at 5000 ng/mL. After six-hour treatments with this dose, marrow was depleted of CFU-Mix, BFU-E, and CFU-GM, but produced CFU-GM in LTMCs, indicating an ongoing input of CFU-GM from a surviving pre-CFU-Mix population. In contrast, elimination of the latter may be inferred from the absence of CFU-GM in LTMCs exposed for three to seven days to diaziquone at only 150 ng/mL. Under these conditions, CFU-F recovery was 40% and adherent stromal layers in LTMCs were similar to untreated controls regarding rate of development and cellular composition. Our in vitro pre-CFU-Mix-ablative regimen correlates with clinical data that show prolonged but reversible myelosuppression at steady-state diaziquone plasma levels of 101 +/- 10 ng/mL (mean +/- standard error of mean) during 7-day constant infusions. IN CONCLUSION hematopoietic cells are more sensitive than marrow stromal cells to the dose- and highly time-dependent cytotoxicity of diaziquone, a direct drug-induced noxious effect on the marrow microenvironment is an unlikely cause of the isolated episodes of marrow necrosis after the use of diaziquone in vivo, and prolonged infusion of diaziquone represents an attractive means for achieving myeloablation in selected clinical situations.