A randomized study of methotrexate vs. 6-mercaptopurine and prednisone was carried out in 45 previously untreated patients with acute leukemia. Methotrexate was given orally in divided daily doses for 5-day courses a t intervals of ten to 14 days. Six-mercaptopurine and prednisone were given daily until toxicity or hematologic response occurred. Patients surviving 6 weeks a n d failing to respond to the first treatment regimen were switched to the second program. Five of 29 patients who received methotrexate either as first or second treatment responded-four with partial remissions a n d one with a complete remission. Eight of 34 patients who received 6-mercaptopurine and prednisone responded-two with complete remissions and six with partial remissions. T h e response rate in both g o u p s of patients was low and there was n o significant difference between the two groups. I n a second unrandomized group of 20 patients treated with methotrexate there were two partial.
responses.
NTIL RECENTLY, N O TREATMENT REGIMEN
U for acute leukemia in adults has produced more than 20% complete remissionsabsence of leukemic cells in the marrow and peripheral blood and disappearance of all clinical evidence of organ infiltration.294,5,10 Generally, 6-mercaptopurine has been regarded as standard therapy. Bridges et al.3 found the combination of 6-mercaptopurine and prednisone somewhat more effective than 6-mercaptopurine alone. Huguley et a1.6 reported that eight of 24 patients treated with small oral divided daily doses of methotrexate in 5-day courses repeated at intervals of ten to 14 days achieved complete remissions.
The Southeastern Cancer Chemotherapy Cooperative Study Group studied two therapeutic regimens in adults with acute leukemia to follow-up the observations of Huguley et a1.6 The effectiveness of daily divided doses of methotrexate in short courses was compared with that of daily administration of 6-mercaptopurine and prednisone.
EXPERIMENTAL DESIGN
Selection of Patients
Category A : Patients who met the following criteria were randomized between the two regimens under study.