Analysis of Survival Gains in Myelodysplastic Syndromes after a Decade of the Modern Treatment Paradigm

Abstract

Introduction:The treatment of myelodysplastic syndromes (MDS) changed after the approval of lenalidomide in 2005, and 2 hypomethylating agents (HMA): azacitidine in 2004 and decitabine in 2006. Erythropoietin-stimulating agents (ESA) and reduced intensity conditioning stem cell transplantation (RIC-SCT) also became more widely utilized around that period. Analyses evaluating potential overall survival (OS) improvements in MDS since 2007 across different cancer registries have shown conflicting results. ‘Real-world’ survival trends are important to measure the true impact of these treatments outside clinical trials, especially as MDS treatment innovations may not reach all patients. We assessed the temporal change in OS and cancer-specific survival (CSS) of MDS patients in the US over the decade before (2001 - 2006) and after modern treatment options were established (2007 - 2016).Methods:Adult subjects diagnosed with MDS between 2001 and 2016 were identified in Surveillance, Epidemiology, and End Results (SEER), and categorized in 2 groups based on year of diagnosis: 2001 - 2006 (cohort 1) and 2007 - 2016 (cohort 2). MDS histologic risk was classified into low, intermediate, and high, using International Classification of Diseases for Oncology 3 rd Edition (ICD-O-3) codes. Cause of death (COD) reported to State registries (SEER recode) was used to determine CSS, defined as death from MDS or acute myeloid leukemia. Only cases with histologic confirmation of ICD-O-3 codes and complete follow up records were analyzed. Kaplan-Meier estimation was used to summarize unadjusted OS distribution. To assess the association of cohort 2 with OS and CSS gains, follow up duration was restricted to a maximum of 5 years in both cohorts and survival analysis was performed using multivariable Cox-proportional hazards regression model adjusting for age, sex, race, ethnicity, MDS risk, and geographic location.Results:We included a total of 42,217 patients with MDS, 13,633 (30.8%) in Cohort 1 (2001 - 2006) and 30,584 (69.2%) in Cohort 2 (modern treatment era). Subjects in cohort 2 were slightly older (mean age 73.8 vs 73.2 years, p<0.001) and included more males (58.6% vs 56.3%, p<0.001). Cohort 1 included more subjects with low MDS histologic risk (27.8% vs 17.2%) and fewer subjects with high MDS risk (18.4% vs 20.5%) (Table 1). Median OS for low, intermediate, and high risk MDS were 44 months (95%CI, 41.5 - 46.5), 27 months (95%CI, 25.7 - 28.3) and 10 months (95%CI, 9.3 - 10.7) in cohort 1, and 48 months (95%CI 45.9 - 50.1), 26 months (95%CI 25.3 - 26.7) and 11 months (95%CI 10.6 - 11.4) in cohort 2, but these differences were not significant. In the multivariable model adjusted for age, sex, race, ethnicity, MDS risk, and geographic location, cohort 2 was associated with a significantly lower hazard of overall death compared to cohort 1, HR for OS of 0.97 (95%CI, 0.95 - 0.99, p<0.001). Similarly, the modern era of treatment was associated with lower cancer-specific death compared to cohort 1, HR for CCS 0.93 (95%CI, 0.89 - 0.93, p=0.038). MDS histologic risk was the strongest factor associated with higher risk of overall and cancer-specific death. Other factors significantly associated with worse OS and CSS were advancing age, male sex, Hispanic ethnicity and unmarried status (Table 2). When analysis was restricted to patients with high risk MDS, cohort 2 was associated with a lower hazard of cancer-specific death, HR for CSS 0.90 (95%CI, 0.84 - 0.94, p=0.006), but no significant difference in overall death, HR for OS 0.96 (95%CI, 0.91 - 1.02, p=0.17).Discussion:In a SEER analysis, we found that the modern paradigm of MDS treatment, including access to lenalidomide, HMA, ESA and RIC-SCT, is associated with only modest survival gains at the population level. Across all MDS risk groups, improvements in cancer-specific death were larger than those seen for overall death and, in high risk MDS, a significant gain in CSS did not translate into longer OS. These data suggest there is a need for targeting non-cancer excess mortality in patients with MDS, who usually present with a high comorbidity burden. Strategies to optimize medical conditions coexisting with MDS and better supportive care may help consolidate the gains associated with currently available MDS-directed therapies. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.