PB1922: EFFECTIVENESS OF AZACITIDINE IN FRONTLINE HIGHER-RISK MYELODYSPLASTIC SYNDROME (HR MDS): A RETROSPECTIVE COHORT STUDY OF 382 PATIENTS

Abstract

Background: 5-azacitidine (AZA) is the frontline standard of care in HR MDS, defined as intermediate-, high-, or very-high-risk MDS per the Revised International Prognostic Scoring System (IPSS-R). Real-world studies reporting on the effectiveness of AZA were often limited by sample size or lacked standardized response assessments based on International Working Group (IWG) criteria. Aims: To assess the effectiveness of frontline AZA in treatment-naive adult patients with HR MDS in a real-world setting. Methods: This retrospective cohort study examined outcomes in treatment-naive adult patients with HR MDS who started frontline AZA between 1 Jan 2014 and 1 Jan 2020, identified from the US Oncology Network Electronic Health Record (EHR) database. Key exclusionary criteria were receipt of any prior antileukemic therapy, stem cell transplant within 6 months of AZA initiation, active second malignancies, participation in a clinical trial during the study period, or off-label use of AZA regimens. Data were abstracted by targeted chart review into an electronic case report form from AZA initiation until the earlier of 1 Jan 2021 or death. Effectiveness of AZA was assessed using IWG 2006. Real-world equivalent (based on EHR only) of these endpoints were also derived. All responses were adjudicated by an independent hematologist. Results: Overall, 932 patients with HR MDS were identified, of whom 382 were eligible. Study population was majority male (243 [63.6%]), had a median age of 74 years, and consisted of IPSS-R intermediate-, high-, and very-high-risk categories in 164 (42.9%), 150 (39.3%), and 68 patients (17.8%), respectively. There were 101 patients (26.4%) with poor cytogenetic risk and 46 (12%) with TP53 mutation. Overall median duration on the study was 12.9 months, and median duration on AZA was 4.9 months; reasons for AZA discontinuation were: disease progression, 135 (35.3%); adverse event, 55 (14.4%); death, 41 (10.7%); and palliation, 36 (9.4%). Based on IWG 2006, 30 patients (7.9%) achieved complete remission (CR), with 240 (62.8%) achieving an objective response (OR; defined as CR, marrow CR [mCR], partial remission [PR], or hematologic improvement [HI]). IWG 2006–based median (95% CI) duration of OR, progression-free survival, time to next treatment, and overall survival (OS) were 9.5 (7.7-10.6), 9.1 (8.0-10.8), 11.5 (9.5-12.6), and 17.9 (15.5-21.7) months, respectively. Presence of poor cytogenetic risk appeared to reduce OS (Table). Estimates for response categories based on IWG 2006 differed from those based on EHR, but OR rates (ORRs) were similar (Table). Table - All patientsIWG 2006(n=382) All patientsEHR based(n=382) Poor cytogenetic riskIWG 2006(n=101) CR, n (%) 30 (7.9) 14 (3.7) 8 (7.9) 95% CI 5.4-11.0 2.0-6.1 3.5-15.0 mCR, n (%) 31 (8.1) 13 (3.4) 10 (9.9) PR, n (%) 0 120 (31.4) 0 HI, n (%) 179 (46.9) 112 (29.3) 43 (42.6) ORR, n (%) 240 (62.8) 259 (67.8) 61 (60.4) 95% CI 57.8-67.7 62.9-72.5 50.2-70.0 Duration of CR, median, months 12.0 13.5 15.6 95% CI 7.7-15.6 4.5-21.5 1.7-39.2 OS, median, months 17.9 12.8 95% CI 15.5-21.7 10.3-15.5 Summary/Conclusion: In this, one of the largest real-world HR MDS cohorts, CR and OS with AZA in treatment-naive patients were limited. Novel agents are urgently needed to improve the CR and survival rates in these patients.