Frontline azacitidine (AZA) or intensive chemotherapy (ICTx) in older AML patients.

Abstract

6530 Background: AZA is approved in the EU and USA to treat high-risk MDS and AML with < 30% marrow blasts. As its off- label use is increasing, its role in management of older AML patients, especially those with > 30% marrow blasts, must be evaluated. Methods: We compared 124 AML patients (aged 65y+) treated upfront with AZA in the French patient named program (Thepot, ASH 2009) to 403 non CBF-AML patients from the ICTx ALFA-9803 trial (Gardin, Blood 2007). Risk factors including PS, age, WBC, prior MDS, and cytogenetics were analyzed, with overall survival (OS) and 3-month death rate as endpoints. Results: As expected, patients selected to receive AZA had highermedian age (75 vs. 71 years) (p < 0.001), lower median WBC count (3.1 vs 5.0 G/L, p = 0.02), more frequent prior MDS (43% vs. 16%, p < 0.001) and adverse cytogenetics (43% vs. 19%, p < 0.001), than patients who received ICTx, but similar PS (72% 0-1). Despite these differences, our ICTx- derived risk index, based on adverse cytogenetics, PS, WBC count, and age (Malfuson, Haematologica 2008) was also a powerful predictor of OS after AZA (p < 0.001). In patients with a poor-risk index, the 3-month death rate was 38% in both cohorts, compared to only 18% and 14 % in AZA and ICTx patients with a low-risk index, respectively. AZA patients with PS > 1, WBC > 15 G/L (best cut-off for AZA), and adverse cytogenetics did poorly, while post- MDS AML and age had no prognostic impact in this cohort. After restricting the comparison to patients with WBC ≤ 15 G/L, 18 m-OS was 31% (20-43) in the AZA vs. 40% (33-47) in the ICTx cohort (p = 0.22), with similar early death rates (27% vs 23%). No differences in outcome appeared between AZA and ICTx cohorts when focusing on post-MDS AML, normal cytogenetics (18-m OS, 64% [35-82] vs. 49% [39-59], p = 0.27), or in patients with adverse cytogenetics, including those with isolated +8, abn(5q) or-7/del7q (18-m OS, 10% [3-23] vs. 15% [6-28], p = 0.69). Conclusions: This comparison suggests that usual AML and host risk factors defined for ICTx have similar impact after AZA. If one excludes patients with WBC > 15 G/L who did poorly with AZA, outcomes seem equivalent, even in cytogenetically normal AML. A controlled trialis thus warranted to compare AZA and ICTx in those older patients. Author Disclosure Employment or Leadership Position Consultant or Advisory Role Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration Celgene Celgene Celgene