Which AML Subsets Benefit from Leukemic Cell Priming during Chemotherapy? Long-Term Analysis of the ALFA-9802 GM-CSF Study

Sensitization of leukemic cells with G- or GM-CSF may enhance the cytotoxicity of chemotherapy in younger adults with newly diagnosed acute myeloid leukemia (AML). In this multicenter randomized trial conducted by the Acute Leukemia French Association (ALFA), 259 patients (median age, 38 years; range: 15–50) were randomized at baseline to receive a timed-sequential induction (DNR 80mg/m²/d d1-3, AraC 500mg/m²/d d1-3 and 500mg/m²/12h d8-10, Mitox 12mg/m²/d d8-9) either without (135 patients) or with GM-CSF (124 patients) administered i.v. at 5μg/kg/d from day 1 to day 10. Patients reaching remission (eventually after salvage therapy) and not eligible for allogeneic stem cell transplantation were then randomized to compare the timed-sequential ALFA-9000 consolidation (Mitox 12mg/m²/d d1-3, VP16 200mg/m²/d d8-10, AraC 500mg/m²/d d1-3, 8-10) to the CALGB postremission chemotherapy, which includes 4 cycles of high-dose AraC (3g/m²/12h d1,3,5) followed by 4 additional DNR/AraC courses. Patients initially randomized in the GM-CSF arm received GM-CSF concurrently with chemotherapy during all cycles of consolidation therapy. First published results were encouraging showing that priming of leukemic cells with GMCSF is a means of enhancing the efficacy of chemotherapy in younger adults with AML (Thomas et al., Leukemia 2007). The effects of GM-CSF on EFS and according to distinct biological subgroups are reported herein with a longer follow-up.

A significantly higher CR rate was observed after induction chemotherapy in patients primed with GM-CSF (88% for patients receiving GM-CSF versus 78% for those not receiving GM-CSF; p = 0.04). With a median follow-up of 5.4 years, the EFS rate was confirmed to be better in the GM-CSF group (43% vs 34%; p = 0.04). However, the difference between the two arms remained not statistically different in terms of OS. GM-CSF did not improve the outcome in the subgroups with favorable cytogenetics constituted by CBF leukemias (61% at 5 years in the GM-CSF group vs 62% in the non GM-CSF group; p = 0.9). Research in defining prognostic factors has moved from classical cytogenetic to an examination of molecular markers. Molecular markers are particularly important for patients with AML and a normal karyotype, identifying within this group two genotypes, NPM1⁺ FLT3-ITD⁻ and CEBPA⁺ FLT3-ITD⁻, with a relatively favorable outcome which were classified here as intermediate-1. In the present study cohort, these intermediate-1 genotypes were associated with a risk profile comparable with that of CBF AML (5-year EFS at 67%). Comportment after priming with GM-CSF was also comparable showing no difference in terms of long-term EFS compared to absence of priming. Although results were not statistically significant, patients with intermediate-risk cytogenetics other than intermediate-1, referred here as intermediate-2, and patients with unfavorable cytogenetics benefited from GM-CSF therapy (5-year EFS, 24% vs 18%, p = 0.2). The benefit of GM-CSF was in part related to a higher proportion rate of CR after the initial course of induction chemotherapy (89% vs 67%, p = 0.04) for unfavorable cytogenetics and a trend for a lower relapse rate (38% vs 52%) in intermediate-2 patients. It appeared that the difference in terms of EFS probability in this population resulted from GM-CSF-mediated activation of subpopulations of leukemic cells that were initially the more proliferative. Actually, the effect of GM-CSF became significant when considering only patients with higher initial WBC count (> 10 × 10⁹/l) (5-year EFS, 32% vs 5%, p = 0.003), while there was no difference in patients with lower initial WBC count (< 10 × 10⁹/l). Among frequent mutations observed in AML, FLT3-ITD provides a proliferative advantage to a population in which normal maturation is blocked. Similarly, MLL fusion proteins also confer a remarkable growth/survival advantage. When combining these two abnormalities for comparison of the effect of GM-CSF priming, the difference in terms of EFS was highly significant (5-year EFS, 39% with GM-CSF vs 8% without GM-CSF in the population of patients FLT3-ITD or MLL positive, p = 0.005). The results of priming in the ALFA-9802 trial confirm that chemotherapy and sensitization of leukemic cells and their progenitors by GM-CSF is a plausible strategy for improving the outcome of younger patients with newly diagnosed AML particularly for those with intermediate-2 or unfavorable cytogenetics displaying higher proliferative features.