Intravenous 6-mercaptopurine decreases salvage after relapse in childhood acute lymphoblastic leukemia: a report from the Children's Cancer Group study CCG 1922.

Purpose: To compare outcomes of patients with NCI standard risk acute lymphoblastic leukemia (ALL) who relapsed after being randomized to receive either oral or intravenous 6-mercaptopurine (6MP) in the Children's Cancer Group study CCG 1922.

Patients And Methods: CCG 1922 accrued patients from March 1993 to August 1995. A total of 1,060 patients were randomly assigned to four treatment groups: oral 6MP plus prednisone (OP), intravenous 6MP plus prednisone (IP), oral 6MP plus dexamethasone (OD), and intravenous 6MP plus dexamethasone (ID). During the 2nd through 4th month of therapy groups OP and OD were treated with 75 mg/m(2)/day of oral 6MP for 70 days and groups IP and ID with 1,000 mg/m(2)/week of intravenous 6MP over 10 hr for 11 doses. All patients received a single delayed intensification and all received oral 6MP in maintenance.

Results: Patients randomized to oral 6MP had significantly better 5-year overall survival (96 +/- 1% vs. 92 +/- 1%; P = 0.008). There was, however, no statistically significant difference in the event-free survival (EFS). Of the 179 patients who relapsed, 84 had a second or later event and 68 have died. Forty of the 84 second events were a death. Survival after relapse was significantly greater for patients randomized to oral 6MP during consolidation than those receiving intravenous 6MP (P = 0.002, log rank test) with 4-year survival post-relapse of 67 +/- 6% vs. 48 +/- 6%. The steroid randomization had no influence on outcome. Post-relapse therapy details are not available and if different between groups may have influenced the outcome.

Conclusion: Treatment with intravenous 6MP during a brief period of total therapy had a significant negative impact on the prognosis in childhood ALL even though oral 6MP was used during maintenance.