Optimizing therapy for acute myeloid leukemia based on differences in intracellular metabolism of cytosine arabinoside between leukemic blasts and normal mononuclear blood cells.

The increasing insights into the pharmacokinetics and the metabolism of arabinoside C (AraC) have improved the rationale for its application in leukemia therapy and have led to a pharmacologically directed design of antileukemic treatment. The current study aims at adding to this approach by detecting differences in the intracellular metabolism of AraC 5'-triphosphate (AraCTP) between leukemic and normal mononuclear blood cells. Measurements of intracellular AraCTP levels were complemented by determinations of plasma AraC and arabinoside uridine (AraU) concentrations and were performed in 26 patients with acute myeloid leukemia (AML) who were undergoing combination therapy, including high-dose (1.0 or 3.0 g/m2 x 2/day) AraC. Plasma AraC concentrations showed a linear relationship to the applied AraC dose but did not correlate with intracellular AraCTP levels. Substantial differences in AraCTP retention times were revealed, during 3-h infusions of either 1.0 or 3.0 g/m2 AraC in leukemic blasts from 10 patients with t1/2 values of 1.60-7.63 h (median, 2.42 h). In addition, AraCTP levels declined in only one patient by > 10% within the first hour after the end of therapy and remained constant or even increased up to 1.5-fold during a posttreatment period of 1-2.5 h in the other nine cases. In contrast, AraCTP retention times were relatively uniform in normal mononuclear blood cells from 11 patients, with t1/2 values of 3.34-5.29 h (median, 3.85 h). More importantly, AraCTP levels dropped by > 10% within the first hour after the end of the high-dose AraC infusion in eight of 11 cases. A posttherapeutic increase of > 10% was not observed in any patient. These differences in AraCTP pharmacokinetics between leukemic and normal blood cells provided the basis for a modified timing of AraC administration with the aim of selectively maintaining cytotoxic AraCTP levels in leukemic blasts while allowing an intermittent drop of AraCTP levels in normal cells. This modification may result in higher antileukemic activity without increasing the damaging effect on normal cells and may, thus, improve the therapeutic index for AraC.