Secondary (s)AML from MDS is more frequent in older AML patients, and associated with an overall worse outcome with standard chemotherapy than de novo AML, particularly after MDS of longer duration (1). The azanucleoside hypomethylating agents 5-azacytidine (Vidaza) and 5-aza-2′-deoxycytidine (Decitabine, DAC) are active in MDS and, as recently shown, also AML. Compared to other predictors of response to these drugs, MDS duration prior to treatment thus far has received only limited attention, with two recent publications reporting conflicting results (2, 3). To independently validate our finding that shorter duration of MDS prior to DAC treatment may be a novel predictor of poor outcome (2, 4), we now applied this parameter to a large trial of low-dose DAC in AML pts (aged >60 years and judged ineligible for standard induction chemotherapy), about half of them with sAML from MDS with variable disease duration.
Comparisons of response rate (RR, i.e. CR or PR) and overall survival (OS) from start of treatment according to MDS duration (pre-specified categorization according to quartiles) were performed post-hoc in 109 patients (pts) with previously untreated sAML (median age 72 years) treated with DAC (given over 72 hours, every 6 weeks, for up to 4 courses, followed by “maintenance” with 3 daily 1-hour infusions of DAC 20 mg/m2 every 4–6-weeks). Median WBC prior to treatment was 5.200/μl, median serum LDH 279U/l, 31.2% of pts had adverse cytogenetics, 82.6% had a performance status > 1, and 80.7% had a comorbidity index (HCT-CI) >=1. Comparisons by logistic regression and Cox regression (univariate and multivariate, adjusted for other prognostic factors showing an effect in this population of sAML pts) were performed.
Of the 227 AML patients treated within the 00331 trial, 109 (48%) had prior MDS with known MDS duration, with a median duration of 8 (25% quartile 3, 75% quartile 25, range 1–101) mths. The overall RR in these pts was 26/109 (24%), the overall 1 yr OS rate was 31% (94 deaths). A comparison of RR according to MDS duration revealed a trend to an increase in RR with longer duration of MDS [<3: 4/25 (16%), 3–8: 5/29 (17%), 8–25: 7/27 (26%), >=25 mths: 10/28 (36%), test for heterogeneity p=0.29, test for trend p=0.06]. Similarly, when OS from start of DAC was analyzed according to this parameter, for pts with previous MDS of longer duration there was a trend to better outcome [<3: 1 yr OS rate 23%, 3–8: 28%, 8–25: 26%, >=25 mths: 46%, test for heterogeneity p=0.17, test for trend p=0.16]. When these analyses were adjusted for other prognostic factors showing an effect in this population of sAML pts (comorbidity index, sLDH with respect to RR, and performance status, comorbidity index, and white blood count with respect to OS), the results were similar (effect of MDS duration with respect to RR: test for heterogeneity p=0.35, test for trend p=0.06, and effect of MDS duration with respect to OS: test for heterogeneity p=0.04, test for trend p=0.11).
In this large cohort of uniformly treated pts with sAML, MDS of longer duration appeared to be associated with a better outcome, even after adjusting for important other prognostic factors. These results are supported by a similar analysis of MDS pts randomized in the 06011 EORTC intergroup trial (which compares DAC to Best Supportive Care), where MDS patients with longer (>=3 mths) disease duration prior to treatment also had better outcome (4). They warrant application of this discriminator in the evaluation also of other non-intensive AML treatment modalities.
1. Estey et al., Blood 90:2969-77, 1997
2. Wijermans et al., Ann. Hematol. 84 Suppl 1:9-14, 2005
3. Kantarjian et al., Cancer 109:265-73, 2007
4. Lübbert, Suciu et al., Abstract submitted, ASH 2010
Off Label Use: decitabine is FDA-approved for treatment of MDS and AML with up to 30% blasts. In the present study, patients with AML and higher blast percentage were treated. Platzbecker: Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Döhner: Pfizer: Research Funding.
Asterisk with author names denotes non-ASH members.