Introduction: Treatment of the elderly AML patient remains challenging. Standard induction chemotherapy may not be well tolerated and increased rates of higher risk disease have led to poor long-term outcomes. New strategies are needed in this population.

Methods: A retrospective review of consecutive Acute Myelogenous Leukemia patients age >60 years treated with either 5-azacitidine 75mg/m2/d 7 days per month (5-aza) or 3 days idarubicin 12 mg/m2 with 7 days cytarabine 100 mg/m2 continuous infusion (3+7). Patients were offered both treatments, and supportive care alone, and chose their own therapy.

Results: Between November 2004 and August 2006, 33 elderly patients with AML underwent therapy with either 5-aza (n=11, median age = 74 years) or 3+7 (n=22 median age = 67) (p=0.07) at Hackensack University Medical Center. Consolidation therapy in the 3+7 cohort consisted of high-dose cytarabine (n=12) and allogeneic transplantation (n=2). Rates of secondary leukemia were balanced between groups (27% 5-aza and 36% 3+7; p=0.9). SWOG cytogenetic risk categories were high-risk in 36% 5-aza and 45% 3+7, and intermediate risk in 64% 5-aza and 55% 3+7 (p=0.9). Marrow blast counts (median) at initiation of therapy were 42% in 5-aza and 65% in 3+7 (p=0.01). Median survival from diagnosis was similar between both therapies at 397 days for 5-aza and 276 days for 3+7 (Log-rank p=0.7). Patients with high-risk cytogenetics fared poorly in the entire cohort with a median survival of 154 days versus 435 days with intermediate risk cytogenetics (p=0.002). Patients with high-risk cytogenetics did not benefit from 5-aza therapy (median survival 35 days) compared to 3+7 (median survival 214 days) (p=0.09), but the trend was reversed among intermediate-risk cytogenetic patients (5-aza median survival 435 days; 3+7 median survival 276 days; p=0.13). Early mortality (<100 days) occurred in 36% 5-aza (2 infection, 1 disease, 1 cardiac) and 18% 3+7 (2 infection, 2 organ toxicity) (p=0.47). Using IWG MDS criteria for hematologic responses, red cell responses among 5-aza pts were CR:36%, PR:18% and among 3+7 pts were CR:68% (p=0.17 for CR rate and p=0.69 for ORR). Platelet responses among 5-aza were CR:36%, PR:18% and among 3+7 were CR:68% (p=0.17 for CR rate and p=0.69 for ORR rate). Median red cell transfusions from diagnosis were 7 units in 5-aza and 15 units in 3+7 (p=0.03). Median platelet transfusions were 4 units in 5-aza and 11 units in 3+7 (p=0.02). Bacteremia occurred in 2 of 11 (18%) 5-aza patients (3 episodes) and 17 of 22 (77%) 3+7 patients (30 episodes) (p=0.002). Patients receiving 5-aza induction spent a median of 11% of their at-risk days in the hospital compared to 33% median at-risk days for patients receiving 3+7 (p=0.12). Overall for the entire cohorts, 5-aza patients spent 213 days in hospital among 2651 at risk days (8%) versus 1229 days in hospital among 3759 at risk days (33%) (p=0.001).

Conclusions: In this series of elderly (>60 year old) patients with AML, 5-azacitidine yielded similar survival outcomes compared to standard 3+7 induction and was associated with significantly less transfusional support, bacteremias, and hospital days. Although not a curative therapy, poor outcomes with standard chemotherapy in the elderly AML patient make consideration of the well tolerated outpatient 5-azacitidine therapy attractive.

Disclosures: 5-azacitidine is approved for myelodysplastic syndromes. Study involves use of 5-azacitidine in acute myelogenous leukemia (off-label).

Author notes


Corresponding author