Comparative Effectiveness of Azacitidine Versus Decitabine Among Older Adults Diagnosed with Higher-Risk Myelodysplastic Syndromes (HR-MDS)

Background: The hypomethylating agents (HMAs) azacitidine and decitabine are both approved for treatment of MDS, but only azacitidine has been shown to have an overall survival (OS) advantage over conventional care in a randomized phase III trial for patients with HR-MDS. Azacitidine has not been directly compared to decitabine in clinical trials of patients with HR-MDS. In addition, it is unknown whether the OS observed with azacitidine in clinical trials would hold in “real-world” clinical practice.

Methods: In a retrospective cohort study, patients were identified from the Surveillance, Epidemiology, and End Results-Medicare database using the following criteria: 1) having an International Classification of Diseases for Oncology, 3^rd edition (ICD-O-3) code for MDS; 2) diagnosed between 1/1/2005 and 12/31/2011; 3) aged ≥66 years at diagnosis; and 4) enrolled in Medicare Parts A and B continuously through death or end of study (12/31/2013). Patients with the ICD-O-3 code 9983 (RA with excess blasts) were classified as having HR-MDS. Healthcare Common Procedural Coding System codes identified outpatient treatment with azacitidine and decitabine. Patients who had <10 days of treatment with azacitidine or decitabine, and patients who received both HMAs were excluded. We used Medicare claims to calculate a modified Elixhauser comorbidity count and disability status. Median household income and percentage of adults with high school education or less were derived from census data. Kaplan–Meier analysis was used to estimate OS probabilities from date of HMA initiation until death or end of study. Multivariate Cox proportional hazard models were utilized to analyze covariate-adjusted survival and estimate hazard ratios. The covariates included year of diagnosis, age at diagnosis (66-69, 70-74, 75-79, and 80+ years), gender, race, % of adults with high school education or less, median household income, the modified Elixhauser comorbidity count and disability status score, time from MDS diagnosis to HMA initiation, and pre-HMA use of erythropoiesis-stimulating agents. 

Results: Of 27,422 patients diagnosed with MDS, 2,025 were eligible for inclusion; 1,580 (78%) received azacitidine and 445 (22%) received decitabine. Median age at diagnosis was 77 years, 64% were males, and 91% were white. Patients received a median of 6 cycles. A total of 1,205 patients (60%) received ≥4 cycles of either HMA. By end of follow-up, 1,729 patients (85%) had died. Median OS for the entire cohort from time of HMA initiation was 15 months (95% Confidence Interval [CI]: 14-15 months). Of the 532 HR-MDS patients who received HMAs, 76% received azacitidine and 24% received decitabine. The median OS for these 532 patients was 12 (95% CI: 11-14) months, and there was no statistical difference in median OS between patients who received different HMAs; 11 (95% CI: 10-14) months for azacitidine vs. 12 (95% CI: 11-16) months for decitabine, p = 0.26, Figure. In multivariate analysis of OS among patients with HR-MDS, the hazard ratio was 0.99 (95% CI: 0.78-1.24) indicating no significant improvement (or decrement) for decitabine compared to azacitidine.

Conclusions: In this large cohort of HR-MDS patients treated with HMAs, we found no significant differences in OS between azacitidine versus decitabine. Additionally, we found significantly shorter OS for HR-MDS patients treated with azacitidine in real-world clinical practice compared to results from randomized trials (11 vs. 24.5 months). Although the shortened OS could be partly related to the older age of this patient cohort, our data provides a real-life reference for counseling older MDS patients regarding this drug's potential benefits. Additionally, these results provide rationale for older patients with HR-MDS to consider enrollment into clinical trials upfront, potentially in HMA-based combinations, rather than defaulting to the use of “standard-of-care” HMA monotherapy. 

 

Figure. Kaplan-Meier survival curves for HR-MDS patients (n = 523)