Metabolism of ara-C by blast cells from patients with ANLL.

The dose-response relationship between extracellular concentration of cytosine arabinoside (ara-C) and intracellular formation of the putative active metabolites of ara-C [ara-C incorporation into DNA and intracellular pools of ara-C in triphosphate form (ara-CTP)] was investigated in blast cells obtained from patients with acute nonlymphocytic leukemia (ANLL) by exposing these cells in vitro to 10, 100, or 1,000 nmol/L of ara-C. We studied 23 untreated patients who subsequently achieved complete remission (CR) with a regimen using daunorubicin and conventional doses of ara-C (ara-C-sensitive group), and 30 patients judged to be ara-C-resistant either by failing initial induction therapy (16 patients) or by having relapsed on an ara-C-containing maintenance regimen (14 patients). In both patient groups, ara-C incorporation into DNA and intracellular ara-CTP both displayed statistically significant increases in response to increasing extracellular concentrations of ara-C (P = .0001 in both cases), with the rate of increase of ara-CTP greater than that of ara-C incorporation. Moreover, blast cells from all patients, even those who were most clinically resistant to ara-C, were able to form ara-CTP and to incorporate ara-C into DNA. Each tenfold increment in extracellular ara-C concentration caused an 8.5-fold increase in ara-CTP, but only a 3.6-fold increase in ara-C incorporation into DNA. Thus, the efficiency of incorporation of ara-C into DNA (defined as the ratio of ara-C incorporation to ara-CTP pools) decreased by 58% with each tenfold increment in the extracellular concentration of ara-C (P less than .0001), presumably as a result of the inhibitory effect of ara-CTP on DNA polymerase. Using an analysis of covariance, modest differences were found in the levels of the ara-C metabolite variables in the ara-C-sensitive group as compared with the resistant group. However, because there was considerable overlap in ara-C metabolite formation among the patient groups, it was not possible to predict clinical outcome by these in vitro assessments of ara-C metabolism.