Modulation of 1-beta-D-arabinofuranosylcytosine metabolism by thymidine in human acute leukemia.

Twenty-seven patients with acute leukemia have been treated by sequential 6-day courses of thymidine (30 g/m2 by i.v. continuous infusion, days 1 and 4) and 1-beta-D-arabinofuranosylcytosine (ara-C) (200 mg/m2 by i.v. continuous infusion, days 2,3,5, and 6). Of 25 evaluable patients 4 achieved a complete remission: one of 9 for acute myelogenous leukemia; and 3 of 14 in the blastic crisis of chronic myelocytic leukemia. Six minor responses were also observed. Toxicity was mainly hematological and did not appear to be higher than that expected from ara-C alone. However, thymidine infusions gave rise to headache and somnolence. The clinical benefit of such treatment seems to be limited to the blastic crisis of chronic myelocytic leukemia. Parallel cytokinetic and biochemical studies were performed in order to assess the cytokinetic and metabolic changes induced by both drugs and to correlate them with the clinical response. Recruitment of cells into the S-phase fraction was observed following the first thymidine infusion in the two complete responders and in three of the five nonresponders studied. In contrast to this high pretherapeutic levels of S-phase fraction were observed in most minor responders and in some nonresponders with further decrease following the thymidine infusion. Recruitment of cells into S phase therefore appeared to be an important but not sufficient factor for prediction of complete response to ara-C. Responders in contrast to most nonresponders were characterized by a higher intracellular level of ara-C and its metabolites following the first 24-h infusion of the drug. Deoxythymidine triphosphate and deoxycytidine triphosphate pools were also measured before and during treatment in order to assess if nucleotide pool variations induced by the administration of thymidine can in fact correlate with the intracellular alteration in ara-C metabolism and with clinical response. The level of deoxycytidine triphosphate pools before treatment showed marked interpatient variations but did not correlate with response. As expected, thymidine infusion induced a rise in the deoxythymidine triphosphate pool and a decrease in deoxycytidine triphosphate. The pools, however, generally returned promptly to the pretherapeutic level 24 h after the end of the infusion of thymidine. There were no significant differences between responders and nonresponders in the modulation of these pools.