Survival advantage with decitabine versus intensive chemotherapy in patients with higher risk myelodysplastic syndrome

Comparison with historical experience




Decitabine, a hypomethylating agent, is active and has been approved for the treatment of myelodysplastic syndrome (MDS) and chronic myelomonocytic leukemia. Intensive chemotherapy is an accepted form of therapy for patients with higher risk MDS. The comparative efficacy of these 2 forms of treatment in MDS is unknown. The objective of the current study was to compare the efficacy and toxicity profiles of decitabine and intensive chemotherapy in MDS.


The authors compared lower intensity decitabine therapy (n = 115 patients) with intensive chemotherapy (as it is used in acute myeloid leukemia [AML]) in patients with higher risk MDS. Two comparisons were made with a cohort of 376 historic patients (from 1995 to 2005): The first comparison included a subcohort of 115 patients (Group A) who matched the 115 decitabine study patients according to age, International Prognostic Scoring System, and cytogenetics; and the second comparison included the whole cohort of 376 patients without matching (Group B). A multivariate analysis was performed for outcome.


The complete remission (CR) rate according to AML criteria was 43% with decitabine, 46% with intensive chemotherapy in Group A, and 52% with intensive chemotherapy in Group B. Compared with Group A, mortality at 6 weeks was 3% with decitabine versus 13% with intensive chemotherapy (P = .006) and, at 3 months, 7% with decitabine versus 23% with intensive chemotherapy (P = .001). Survival was better with decitabine versus intensive chemotherapy in Group A (median survival: 22 months vs 12 months; P < .001). A multivariate analysis of survival in all 491 patients who received decitabine or intensive chemotherapy (Group B) selected decitabine as an independent, favorable prognostic factor for survival (P = .006; hazard ratio, 0.74) after accounting for the independent prognostic effect of pretreatment factors.


In this analysis, decitabine was associated with a survival advantage compared with intensive chemotherapy in patients with higher risk MDS. Future studies should evaluate prospectively the results of decitabine versus intensive chemotherapy in this setting. Cancer 2007 © 2007 American Cancer Society.

Myelodysplastic syndromes (MDS) are heterogeneous disorders characterized by a hypercellular bone marrow with dysplastic changes and peripheral cytopenias.1, 2 Several prognostic models, including the International Prognostic Scoring System (IPSS), have been proposed to account for the heterogeneity in MDS.3–5 Prognostic factors include the percent of bone marrow blasts, cytogenetic abnormalities, the degree and number of cytopenias (according to the IPSS), and others.6 Patients with higher risk MDS, usually with ≥5% blasts or in the IPSS intermediate-high risk group, have a poor survival.5 Standard therapies in higher risk MDS include intensive chemotherapy and allogeneic stem cell transplantation (SCT).7–9 Because the median age of patients with MDS is from 65 years to 70 years, allogeneic SCT is possible in only a minority of patients. Intensive chemotherapy may result in complete remission (CR) rates of 40% to 60% but is associated with considerable morbidity and with mortality rates of 20% to 40%, depending on the definition of induction mortality (eg, at 4 weeks vs 8 weeks). Hypomethylating agents like decitabine and 5-azacytidine have shown positive results in MDS and chronic myelomonocytic leukemia (CMML).10–12 They are considered lower intensity therapies and have been associated with low treatment-related mortality as a consequence of myelosuppression.

The long-term prognosis in MDS and acute myeloid leukemia (AML) has been related to improvements in rates of complete remission (CR). However, prognosis also is related to the composite effects of a treatment's capacity to induce CR and its adverse effect on mortality. Responses other than CR may improve prognosis through the induction of quality responses that improve cytopenias.

To our knowledge, no studies to date have compared the long-term results of lower intensity chemotherapy (eg, hypomethylating agents) with results from AML-type intensive chemotherapy in higher risk MDS. This is the focus of the current analysis.


Decitabine Study Group and Therapy

Adults with a diagnosis of MDS or CMML who were referred to The University of Texas M. D. Anderson Cancer Center from November 2003 through July 2006 were offered decitabine therapy on study (Protocol ID03-0180) after informed consent was obtained according to institutional guidelines. Eligibility criteria included 1) age ≥16 years, 2) a diagnosis of MDS (IPSS intermediate- or high-risk) or CMML, and 3) normal organ function, including creatinine ≤2 mg/dL and bilirubin ≤2 mg/dL. Patients who had received prior intensive chemotherapy with cytarabine ≥1 g/m2 were not eligible. The diagnosis of CMML was based on the typical morphologic picture, including unexplained leukocytosis >12 × 109/L that lasted for ≥3 months, excluding other myeloproliferative disorders, and the presence of ≥109/L monocytes.

Patients were randomized to receive decitabine on 1 of 3 schedules: 1) 20 mg/m2 intravenously over 1 hour daily for 5 days; 2) 20 mg/m2 daily for 5 days given in 2 subcutaneous doses daily; or 3) 10 mg/m2 intravenously over 1 hour daily for 10 days. All patients received the same decitabine total dose per course, 100 mg/m2. Details of therapy, monitoring, dose modifications, antibiotics prophylaxis, growth factors support, other supportive care measures, statistical design, and results of the Bayesian randomization have been published in detail.12, 13 The Bayesian design selected decitabine 20 mg/m2 intravenously daily for 5 days as the schedule associated with the highest likelihood of achieving CR, as reported previously.12

The study group has accrued a total of 115 patients. The characteristics of these patients are shown in Table 1.

Table 1. Characteristics of the Decitabine Study Group, the Historic Matched Control Group That Received Intensive Chemotherapy (Group A), and the Overall Historic Intensive Chemotherapy Group (Group B)*
CharacteristicNo. of patients (%)PNo. of patients (%)P
DecitabineGroup AGroup B
  • IPSS indicates International Prognostic Scoring System; MDS, myelodysplastic syndrome.

  • *

    The P values refer to differences between the decitabine group and either of the 2 historic control groups.

No. of patients115115 376 
 ≥60 y80 (70)77 (67).78224 (65).42
 Median [range], y64 [37–89]65 [41–83]64 [35–84]
IPSS risk
 Intermediate 119 (16)18 (16).9834 (9)<.01
 Intermediate 234 (30)33 (29)114 (30)
 High17 (15)19 (16)140 (37)
 Not categorized45 (39)45 (39)88 (23)
Karyotype prior to therapy
 Chromosome 5 and/or 7 abnormalities33 (29)34 (30).66143 (38) 
 Diploid46 (40)51 (44)147 (39).10
 Other36 (31)30 (26)86 (23)
Duration of MDS
 ≥ 6 mo42 (37)43 (37).89150 (40).59
 Median [range], mo3 [0–102]3 [0–120]3 [0–144]
Percent bone marrow blasts
 5–1041 (36)49 (43)<.01114 (30)<.01
 11–1943 (37)39 (34)144 (38)
 20–297 (6)27 (23)118 (31)
 Median [range]. %9 [0–26]13 [5–29]15 [5–29]
 <10 g/dL87 (76)93 (81).42322 (86).02
 Median [range], g/dL9.2 [4.3–13.0)]8.2 [4.0–13.9]7.8 [2.2–15.1]
 <50 ×109/L54 (47)55 (48).89214 (57).08
 Median [range], ×109/L56 [4.5–419]52 [7–270]42 [3–307]
 <1.5 ×109/L71 (62)68 (59).79225 (60).80
 Median [range], ×109/L1.05 [0.11–61.58]1.11 [1.01–56.39]1.00 [0–56.39]

Historic Control Group Treated with Intensive Chemotherapy

To evaluate the results of decitabine versus intensive chemotherapy, we analyzed the details of patients with higher risk MDS who had received intensive AML-type chemotherapy during a reasonably comparable period, ie, since 1995. We excluded from the historic control group patients who did not fit the decitabine protocol eligibility criteria (performance status of 3 or 4, creatinine >2.0 mg/dL, and bilirubin >2.0 mg/dL). In total, 376 historic controls were identified. These patients received AML-type chemotherapy according to the frontline AML study of the period, as detailed previously.7 The regimens used were anthracyclines plus high-dose cytarabine (156 patients), topotecan plus high-dose cytarabine with or without cyclophosphamide (138 patients), clofarabine or fludarabine plus high-dose cytarabine (36 patients), other cytarabine regimens (18 patients), and noncytarabine regimens (28 patients).7 Our previous analysis indicated that these regimens produced equivalent results.7

The comparisons of the decitabine study group were performed in 2 different analyses. In the first analysis, we matched each patient who received decitabine to 1 patient from the historic control group who received intensive chemotherapy. Patients were matched for 1) age (±5 years), 2) chromosomal abnormalities, and 3) IPSS risk group prior to therapy. This historic control group (Group A) included 115 patients. One hundred eleven patients in the decitabine group and 111 patients in the historic control group were matched completely on those 3 criteria, and the other 8 patients were matched as best as possible (with 2 of 3 features matching). In the second analysis, we compared the decitabine study group with the whole cohort of 376 historic patients with higher risk MDS who received intensive chemotherapy in 1995 (Group B).

Response Criteria and Statistical Considerations

To allow comparisons of the decitabine results with other agents/regimens, the response rate of the decitabine trial were updated using the International Working Group (IWG) modified response criteria.14, 15 However, when comparing CR rates with the historic experience, the AML response criteria were used, because these were the only responses coded for the historic control group that received intensive chemotherapy.16 Survival was dated from the start of therapy.


The characteristics of the decitabine study group are detailed in Table 1 and are compared with the 2 historic intensive chemotherapy groups. Because of the selection of historic Group A, as expected, the characteristics of the 2 groups were similar, except that a higher percentage of patients in the historic group had >10% blasts (57% vs 43%; P = .05). Patients who received receiving decitabine therapy had comparable CR rates according to the AML criteria (43% vs 46% for the control group; P = .69) but had a significantly lower 6-week mortality rate (3% vs 13%; P = .006) and 3-month mortality rate (7% vs 23%; P = .001). Survival was significantly better with decitabine overall (median survival, 22 months vs 12 months; 2-year survival, 47% vs 24%; P < .001) (Fig. 1A) and according to age groups (ages <60 years and ≥60 years). In younger patients (aged <60 years), the 2-year survival rates were 63% versus 32% (P = .02). In older patients (aged ≥60 years), the 2-year survival rates were 40% versus 20% (P = .001).

Figure 1.

Survival with decitabine versus a matched historic intensive chemotherapy group (A) and the total historic intensive chemotherapy group (B).

In the second analysis, several pretreatment characteristics of the patients in decitabine study group and in historic control Group B differed (Table 1). According to the AML criteria, the CR rates were 43% versus 52%, respectively (P = .09), and the respective 6-week mortality rates were 3% versus 12% (P = .002). Survival was significantly better with decitabine overall (2-year survival, 47% vs 21%; P < .0001) (Fig. 1B) in older patients (2-year survival, 40% vs 17%; P < .001) and in younger patients (2-year survival, 63% vs 28%; P = .002) (Fig. 2A,B). A multivariate analysis that included pretreatment characteristics selected the following as independent poor prognostic factors: chromosome 5 or 7 abnormalities (P < .01), thrombocytopenia (P < .01), and older age (P < .01). After accounting for these characteristics, therapy with decitabine remained an independent favorable prognostic factor for survival (P = .006; hazard ratio, 0.74).

Figure 2.

Survival with decitabine versus total the historic control group by age ≥60 years (A) and age <60 years (B).

Overall Response to Decitabine by the IWG Modified Criteria

Response to decitabine among the 115 treated patients according to the IWG modified criteria was a complete response in 40 patients (35%), a partial response in 2 patients (2%), a bone marrow complete response in 12 patients (10%), a bone marrow complete response and hematologic improvements in 14 patients (12%), and hematologic improvements in 12 patients (10%; 1 lineage improvement in 8 patients; 2 or 3 lineage improvements in 4 patients).

Among 87 patients with red cell transfusion dependence or hemoglobin levels <10 g/dL, 40 patients (46%) improved their hemoglobin levels by ≥1.0 g/dL or became transfusion independent. Among 83 patients with pretreatment platelet counts <100 × 109/L, 41 patients (49%) ultimately achieved platelet counts ≥100 × 109/L. According to the IWG response criteria, 50 of 83 patients (60%) with pretreatment platelet counts <100 × 109/L had a platelet response, with either an increase from <20 × 109/L to >20 × 109/L and an increase ≥100% (10 of 19 patients; 53%) or an absolute increase in the platelet count ≥30 × 109/L (40 of 64 patients; 62%) that lasted for ≥2 months. Cytogenetic abnormalities were present pretreatment in 69 of 115 patients who received decitabine. A complete cytogenetic response was observed in 24 of 69 patients (35%), and a partial cytogenetic response was observed in 11 additional patients (16%).


In this study, lower intensity therapy with decitabine was associated with better outcome compared with intensive chemotherapy in patients with higher risk MDS. Although the CR rate was lower with decitabine, the early (6-week) mortality rates (3% vs 13%; P = .002) and the 3-month mortality rates (7% vs 23%; P = .001) were significantly lower. This resulted in a significantly better overall survival with decitabine in a comparison with a matched historic control group that received intensive chemotherapy (Fig. 1) and in a multivariate analysis using a larger historic control group that received intensive chemotherapy in the modern era (hazard ratio for decitabine vs intensive chemotherapy, 0.74; P = .006). The survival benefit was not confined to older patients (possibly reflecting a high mortality with intensive chemotherapy) but was also observed in younger patients. This may indicate a role for decitabine maintenance after intensive chemotherapy in younger patients with higher risk MDS or AML, as part of noncross-resistant sequential combination strategies.

To provide comparisons with other recent anti-MDS strategies, we updated the decitabine results in 115 patients treated so far by using the IWG modified response criteria. Overall, 35% of patients achieved CR, and 70% had an objective response. It is noteworthy that significant cytopenias were improved in approximately 50% to 60% of patients with MDS. These results compared favorably with those reported with either decitabine (CR rate, 9%) or 5-azacitidine (CR rate, 7%) in previous randomized trials.10, 11 This may reflect the dose intensity of the current decitabine approach versus the decitabine randomized study. Because of major differences in study groups, the comparative efficacy of decitabine versus 5-azacitidine needs to be evaluated in a randomized trial for patients with MDS.

In summary, the results from the current analysis suggest a survival benefit of decitabine as lower intensity hypomethylating therapy compared with AML-type intensive chemotherapy in patients with higher risk MDS. Future studies should evaluate prospectively the results of decitabine versus intensive chemotherapy in MDS.