In order to determine whether the use of a sequence of chemotherapeutic regimens plus BCG could produce longer durations of remission in adult acute nonlymphocytic leukemia than maintenance therapy with cytosine arabinoside, 6-thioguanine, plus BCG, a randomized study was performed at Washington University. Upon achieving complete remissions with daunorubicin plus cytosine arabinoside, 14 patients were randomized to receive either: Regimen A -cytosine arabinoside, 6-thioguanine, plus BCG each month; or regimen B-sequential regimens consisting of: 1) azacytidine daily for five days; 2) cyclophosphamide plus cytosine arabinoside daily for four days, prednisone daily for five days, plus vincristine on the first day; 3) prednisone, 6-mercaptopurine, and methotrexate daily for five days plus vincristine on the first day; and 4) cytosine arabinoside, 6-thioguanine, plus BCG. Each of the sequential regimens was given during consecutive months, and the cycle was then repeated starting with the first regimen. Median duration of complete remission was 27 months for 8 patients randomized to receive Regimen A, compared to only seven months for 6 patients receiving Regimen B (P < 0.05). The median survival time of patients on Regimen B was only 14 months, and has not yet been reached in Regimen A. At 40 months after diagnosis, 75% of patients on Regimen A remain alive (P < 0.05). Toxicity was equal for the maintenance regimens. Therefore, maintenance therapy with cytosine arabinoside, 6-thioguanine, plus BCG may he superior to the sequence of chemotherapy regimens plus BCG which was employed.
Cancer 46:22-28, 1980.
OLLOWING THE DRAMATIC ADVANCES in chemo-F therapy in acute lymphocytic leukemia during the 1960s, it was logical to expect that chemotherapy of adult acute nonlymphocytic leukemia (ANLL) would also improve. In experimental murine L1210 leukemia, it was observed that cytosine arabinoside and daunorubicin had synergistic antileukemic effects without increased toxicity to normal hematopoeitic stem cells.s As a clinical correlate of this laboratory observation,