Background: For elderly patients unfit for intensive chemotherapy in acute myeloid leukemia (AML), low-dose cytarabine (LDAC; 20 mg SQ BID for 10 days) still remains to be the standard treatment, despite its unsatisfactory complete response (CR) rate of 18% and median overall survival (OS) of < 6 months (Burnett, 2007). Recently, there have been huge efforts to develop more effective and less-toxic therapies, such as decitabine, azacitidine, clofarabine, or gemtuzumab ozogamicin, but their benefits were not concrete, even though they were compared to the classical LDAC. To improve outcomes of the classical LDAC, we modified it by giving a higher dose of cytarabine for an extended duration in combination with oral etoposide. Herein, we present the results.

Methods: Between 2002 and 2014, 93 consecutive older (≥ 60 years) patients with AML, who were unfit for intensive chemotherapy, received 1st cycle of modified LDAC (mLDAC) regimen consisting of cytarabine (20 mg/m2 SQ BID) and oral etoposide (50 mg PO BID) for 14 days. Thereafter, they received additional subsequent cycles (for a maximum of 7 cycles) for 10 days every 6 to 8 weeks. We retrospectively analyzed their overall response (OR), disease-free survival (DFS), and overall survival (OS) rates. In this analysis, OR was defined as CR plus CR with incomplete platelet recovery (CRp) or blood count recovery (CRi).

Results: The median age of patients in our cohort, including 69 (74.2%) with poor performance status (ECOG ≥ 2), 15 (16.1%) with AML with myelodysplastic-related changes or secondary AML, and 13 (14.0%) with poor cytogenetic risk, was 68 years (range, 60-83). The median number of mLDAC regimen cycles which they received was 2 (range, 1-8). Clinically relevant toxicities of grade III-IV including nausea/vomiting, diarrhea, hyperbilirubinemia and neutropenic fever were observed in 4 (4.3%) patients, 6 (6.5%), 3 (3.2%), and 42 (45.2%), respectively, which were comparable with those of classical LDAC (Burnett, 2007). The early mortality rates at 30 and 60 days were 11.8% and 15.0%, respectively. The OR was observed in 45 (48.4%) patients, including 34 (36.6%) CR, 7 (7.5%) CRp, and 4 (4.3%) CRi, within two cycles of mLDAC. With median follow-up duration of 26.1 months, the median DFS and OS were 6.2 and 15.8 months, respectively. For patients who achieved OR, they were 14.5 and 36.9 months, respectively. The OR of patients who had poor cytogenetic risk was not significantly different compared to others (57.1%, 46.2%, and 38.5% for favorable, intermediate, and poor cytogenetics, respectively; P=0.50). However, they showed significantly shorter median DFS (9.8, 6.6, and 5.1 months, respectively; P=0.01) and OS (NR, 1.4, and 5.1 months, respectively; P=0.01) with significantly shorter OR duration (30.6, 19.1, and 8.6 months, respectively; P=0.01). Between 2009 and 2014, among 17 patients treated with hypomethylating agents (HMA; 14 decitabine and 3 azacitidine), 1 CR and 3 partial response were achieved with a median survival of 5.5 months, and 5 patients after HMA treatment failure received subsequent mLDAC, and 3 achieved additional CR (n=2) and CRp (n=1).

Conclusions: These results suggest that the outcomes of classical LDAC in elderly patients with AML can be improved by modifying it, with improved response and survival rates without increasing toxicities, even in patients with poor cytogenetics. Additionally, mLDAC could induce clinical responses in patients with HMA failure. Our mLDAC regimen may become another therapeutic option with emerging novel agents for elderly patients with AML, and these should be confirmed by large randomized trials.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.