High Dose (HD-AraC) Vs Standard Dose Cytosine Arabinoside (SDAraC) during Induction and Value of IL-2 during Maintenance in Acute Myelogenous Leukemia (AML): Impact of AraC Dose on Complete Remission Rate and Toxicity (Results on the first 1700 randomized patients of the AML-12 trial of EORTC and GIMEMA Leukemia Groups)

The AML-12 randomized phase III trial of EORTC-LG and GIMEMA assessed the efficacy and toxicity of HD-AraC (3 g/m²/12 hrs for 4 days) combined with daunorubicin (50 mg/sqm for 3 days) and etoposide (50 mg/sqm for 5 days) vs SD-AraC (100 mg/sqm for 10 days) with the same drugs. Patients (pts) in complete remission (CR) received consolidation consisting of AraC (500 mg/sqm/12 hrs for 6 days) and daunorubicin. Subsequently an allogeneic (allo-SCT) or autologous stem cell transplantation (auto-SCT) was planned according to donor availability and age. A 2nd randomization was performed after CR in pts without a donor: auto-SCT followed or not by low dose IL-2. The trial was powered to detect an 8% difference in the 5-yr survival rate; secondary endpoints were response to induction, DFS, toxicity. Randomization was performed centrally; the 1^st randomization was stratified for age, performance status, WBC and center. Intent-to-treat analysis was done. From 9/1999 till 1/2008, 2005 previously untreated AML pts (APL excluded), age<61 years, (891 by EORTCLG and 1114 by GIMEMA) were randomized. Currently for 1700 pts (857 in SD-AraC vs 843 in HD-AraC arm) sufficient information on remission induction response is available. The median follow up is 3 years. After 1 or 2 courses of induction, CR was achieved in 631 (73.6%) pts (SD-AraC group) vs 680 (80.7%) pts (HD-AraC group): p=0.001. Partial remission or resistance was documented in 161 (18.8%) vs 114 (13.5%) pts, and death in induction in 65 (7.6%) vs 49 (5.8%). Induction toxicity profile and grade was similar in the 2 arms except for conjunctivitis grade 3: 0.1% (SD-AraC) vs 4.4% (HD-AraC). Currently in 909 pts cytogenetic information is not yet available and in 791 pts cytogenetic information is known. Comparison of SD-AraC vs HD-AraC according to cytogenetic subgroups is shown in the Table.

In pts who reached CR, the DFS was similar in the 2 treatment groups: the 3-yr DFS rate was 43.4% (SD-Ara-C) vs 44.6% (HD-Ara-C), hazard ratio (HR)=0.96 (p=0.66). A total of 291 vs 313 DFS-events were reported in the 2 treatment groups: 230 (36.5%) vs 236 (34.7%) relapsed and 61 (9.7%) vs 77 (11.3%) died in CR. Among pts who reached CR, 451 pts had an HLA identical sibling and 860 did not or have not been typed. In the first group, 120/200 (60.0%) SD-AraC vs 134/251 (53.4%) HD-AraC pts underwent an allo-SCT. In the 2^nd one, 211/431 (49.0%) SD AraC pts vs 211/429 (49.2%) HD AraC pts underwent an auto-SCT. In pts with a donor, the 3-yr DFS rate was 51.8% (SD-Ara-C) vs 47.6% (HD-Ara-C), HR=1.13, p=0.41, whereas in those without a donor the 3-yr DFS rate was 39.6% (SD-Ara-C) vs 42.9% (HD-Ara-C), HR=0.91, p=0.31. Evaluation of the first 1700 patients of the EORTC-GIMEMA AML-12 trial shows that, with a median follow-up of 3 years, HD-AraC in the induction treatment leads to a significantly higher CR rate than SD-Ara-C without improving the DFS.