Randomized Comparison of Standard Induction with Daunorubicin (DNR) for 3 Days vs Idarubicin (IDA) for 3 or 4 Days in AML pts Aged 50 to 70 and of Maintenance with Interleukin 2. Final Analysis of the ALFA 9801 Study.

Background: For many years the ALFA Group has used high dose DNR, i.e 80 mg/m2/day for 3d as part of the induction regimen for untreated adult AML pts. As the equivalence of DNR and Ida is not known we conducted a phase III study to compare high dose DNR to Ida. We also tested the effect of IL2 for maintenance.

Methods: Newly diagnosed AML, aged 50 to 70 years, were randomized to receive AraC 200 mg/m2/day IV x 7 d with either DNR: 80 mg/m2/d x 3 d (arm 1) or Ida: 12 mg/m2/d x 3 d (arm 2) or x 4 d (arm 3). Pts who failed could receive a salvage course with Mitoxantrone x 2 d and AraC1g/m2 12 hrs x 4d. Response was assessed after induction+/− salvage chemotherapy. CR pts received 2 consolidation courses according to the induction arm with DNR: 80 mg/m2 or Ida:12 mg/m2 for 1 day (first course) or 2 days (2nd course) and AraC:1g/m2/12 hrs x 4 days. Pts in CCR after the two consolidation courses were randomized to receive or not a maintenance immunotherapy with IL2: 5.10⁶/m2 for 5 days monthly for12 months.

Results: From 2001 to 2006, 468 pts were included (median age: 60 years). The 3 treatment arms were well matched for pretreatment characteristics. CR was achieved in 360/468 pts (77%): 109 (70%) pts arm 1, 129 (83%) arm 2, 122 (78%) arm 3 (p=0.02).70 pts, (45%) pts in the DNR arm reached CR after 1 course vs 97 (62%) and 90 (57%) in Ida arms (p= 0.006). Pts in Ida arms experienced more grade 3 or 4 mucositis (p=0.004) but no differences were observed between the 3 arms for duration of hospitalization, time to PMN or plts recovery, incidence of grade 3 or 4 infectious episodes and treatment related mortality. 3 year cumulative incidence of relapse was 69%, 63%, 62% resp in arms 1,2,3 (p=NS). 3 year EFS was 19%, 30%, 26% resp (p=0.06 for arm 1 vs arms 2 and 3). 3 year OS was 31%, 40% 41% resp in arms 1,2,3 (p=NS). Age (< or >60 years), sex, initial WBC counts, initial LDH (nl or >nl), DNR or Ida arms, need for a salvage course were not predictive for relapse, while 2yCIR was 43%, 64%, 77% among respectively fav, intermediate and unfav cytogenetic risk groups resp (p=0.0046). Of the 219 pts alive in CR after consolidations, 161 (73%) were randomized for maintenance. Only 22 of the 77 pts randomized for IL2 completed the 1 year treatment. 32 and 23 pts stopped IL2 resp because of relapse or intolerance. There were no differences in relapse or OS in both maintenance arms.

Conclusion: Ida treatment even when compared to high doses of DNR yields higher CR rate and more CR after one course. This effect translated in slightly better EFS but not better CIR or OS. IL2 maintenance treatment at least as scheduled in this trial was difficult to apply and showed no impact on disease course.