P766: CLINICAL OUTCOMES IN PATIENTS WITH HIGHER-RISK MYELODYSPLASTIC SYNDROMES RECEIVING HYPOMETHYLATING AGENTS: A LARGE POPULATION-BASED ANALYSIS

Abstract

Background: Myelodysplastic syndromes (MDS) are a spectrum of hematopoietic neoplasms characterized by variable degrees of cytopenias and risk of progression to acute myeloid leukemia (AML). Although the hypomethylating agents (HMA) azacitidine (AZA) and decitabine (DEC) are widely used to treat higher-risk (HR)-MDS in the United States (US), questions remain about their longer-term clinical benefits. Aims: To understand the survival outcomes and progression-related events of patients (pts) treated with AZA or DEC for HR-MDS in the US. Methods: Older adults (≥66 years old) diagnosed with refractory anemia with excess blasts (RAEB), an established proxy for HR-MDS, between January 2009 and December 2017, were included from the Surveillance, Epidemiology, and End Results (SEER)-Medicare linked database. The index date was date of HMA initiation. Pts who received AZA or DEC, had ≥12 months (mo) of continuous enrollment prior to the index date, and who did not receive a stem cell transplant (SCT) prior to the index date were included. Overall survival (OS) from the index date to death was estimated using Kaplan–Meier methodology. Baseline platelet and red blood cell (RBC) transfusion dependence (TD) were defined as ≥2 transfusion episodes in the 8 weeks prior to the index date; baseline transfusion use (TU) was defined as one transfusion episode in the 8 weeks prior to the index date. A competing risk analysis was used to assess the incidence of progression to AML, with death as the competing event. Results: Of 973 pts with HR-MDS treated with HMA, 75.8% received AZA and 24.2% received DEC. Median time from diagnosis to HMA initiation was 1.2 mo (interquartile range [IQR]: 0.6–2.6). Median treatment duration was 5.7 mo (IQR: 2.3–11.5) overall (5.8 mo [IQR: 2.6–11.8] for AZA and 5.0 mo [IQR: 1.9–11.0] for DEC). At baseline, 13.8% of pts had platelet TD/TU, 29.9% had RBC-TD and 27.0% had RBC-TU. Only 6.4% of pts received SCT during follow-up. Median follow-up from the index date until death or censoring was 13.8 mo (IQR: 6.2–25.3). Median OS (mOS) was 13.9 mo (95% confidence interval: 12.9–15.0). There was no significant difference in mOS for pts treated with AZA vs DEC (13.4 vs 15.2 mo, respectively; p=0.22). Pts with RBC-TD/TU had shorter mOS vs those with RBC-transfusion independence (TI) (10.7 vs 18.6 mo, respectively; p<0.0001). Similarly, pts with platelet TD/TU had shorter mOS vs those with platelet TI (8.4 vs 14.9 mo, respectively; p<0.0001). There were no statistically significant differences in OS by the route of AZA administration, insurance category, county metropolitan status, or county poverty level. Overall, 38.0% of HMA-treated pts progressed to AML; median time from HMA initiation to AML progression was 8.2 mo (IQR: 3.5–15.1). Cumulative incidences of AML and death were 25.7% and 30.7%, respectively, at Year 1 and 34.3% and 45.8%, respectively, at Year 2, with the rates of progression slowing over time. Summary/Conclusion: Compared with previously published literature from the SEER-Medicare database (Zeidan Br J Haematol 2016; mOS: 11 mo [AZA] and 12 mo [DEC] in pts diagnosed with HR-MDS between 2004 and 2011), pts with HR-MDS treated with HMA have improved mOS, but OS remains substantially worse vs the landmark AZA-001 trial (Fenaux Lancet Oncol 2009; mOS: 24.5 mo). The incidence of death and AML progression was highest in the first year after initiation of therapy. SCT rates remain very low among pts with HR-MDS, further emphasizing the urgent need for novel treatment options for pts with HR-MDS.