Available evidence suggests that in most patients with LR-MDS the risk of death is not related to disease progression but is mainly attributable to non-leukemic death. 2,17 In addition, a proportion of these patients have prolonged survival that precludes the design of clinical trials adopting OS as a primary endpoint. These challenges have resulted in potentially biased assessment of the effectiveness and appropriate use of the available interventions in this patient population. The EUMDS Registry has identified novel meaningful outcome indicators and clinical endpoints, and reliable measures of response to HCI (Figure 4). The results of our analysis indicate that RBCT density is strongly associated with a decreased OS, even at relatively low dose densities. In addition, we observed that an early decrease in platelet count is an independent adverse prognostic indicator in LR-MDS, and combining relative platelet drop and transfusion dependency allows early identification of patients at risk of rapid progression, and may guide early therapeutic interventions, including allogeneic hematopoietic stem cell transplantation or experimental interventions. Taken together, these results indicate that regular RBCT requirement, early platelet count kinetics, and restriction in HRQoL are early independent and meaningful outcome indicators, and reliable measures of effectiveness of therapeutic interventions, evaluated in this set of studies. These findings support the integration of RBCT requirement and HRQoL in the general core outcome sets and in response criteria in patients with LR-MDS, and have important implications for clinical practice and the design of clinical endpoints. Our results strongly support the adoption of freedom from transfusion as a meaningful clinical endpoint in patients with LR-MDS. Anemia is the main determinant of therapeutic intervention in patients with LR-MDS, and ESA are recommended as first-line treatment for patients with symptomatic anemia. 10 The observational studies within the EUMDS Registry showed that the response rate, as well as the capacity of these agents to delay the onset of a regular RBCT need, is most pronounced in RBCT-naïve patients. These results identified early initiation of treatment with ESA as a major treatment response indicator, and indicate that ESA should be recommended in LR-MDS patients with symptomatic anemia before starting regular RBCT. After the onset of RBCT dependency, patients with LR-MDS are prone to long-term accumulation of iron. 1,43 The EUMDS Registry studies provided evidence that elevated LPI levels are associated with reduced survival in RBCT dependent patients, whereas iron chelation therapy normalizes LPI levels. These findings suggest that NTBI and LPI may serve as early indicators of iron toxicity and a means to measure the effectiveness of iron chelation therapy in patients with LR-MDS. However, qualified NTBI and LPI are only currently available in specialized laboratories. 44 Large observational cohorts with detailed clinical and laboratory data, like the EUMDS cohort, are the ideal framework in which to identify well defined MDS subtypes that may benefit from novel targeted treatments. An example of such a subtype is MDS with loss of parts of chromosome 5, namely del5q; these patients have a relatively favorable outcome on lenalidomide treatment. In order to identify homogeneous subsets of patients within MDS, preliminary evidence has suggested that recently identified mutations in splicing factors may recognize distinct disease entities within myeloid neoplasms. 45 Splicing modulators are now in pre-clinical testing, and are very likely to lead to the introduction of effective drugs for specific groups of MDS patients. Luspatercept, a specific inhibitor of growth and differentiation factor-11, a member of the transforming growth factor β superfamily, induced substantial improvement of anemia, especially in patients with ring sideroblasts. 46 Characterization of individual cases by new genetic markers (one of the main objectives of the MDS-RIGHT project) will allow refined classification of patients into biological subgroups that are expected to respond differently to therapeutic interventions to guide discontinuation of those interventions that are less effective or less cost-effective. The main question is whether RCT data and retrospective cohort data in selected tertiary care centers are representative of the 'real world' data of the older patients with LR-MDS in the general population. A careful comparison of the 'real world' data and the RCT data will be needed in order to provide a clear answer to these questions. Meanwhile, the current analyses of data collected over 10 years in the EUMDS Registry provides relevant and important information which could help assess prognosis and response to standard interventions in this older patient group.