Abstract
Myelodysplastic syndromes (MDS) are clonal disorders characterized by the accumulation of complex genomic abnormalities that define disease phenotype, prognosis, and the risk of transformation to acute myeloid leukemia. The clinical manifestations and overall outcomes of MDS are very heterogeneous with an overall survival that can be measured in years for some patients to a few months for others. Prognostic scoring systems are important staging tools that aid physicians in their treatment recommendations and decision-making and can help patients understand their disease trajectory and expectations. Several scoring systems have been developed in MDS with the International Prognostic Scoring System and its revised version, the most widely used systems in clinical practice and trial eligibility. These models and others use mainly clinical variables that are obtained from bone marrow biopsy and peripheral blood measurements. Adding molecular data to current models may improve its predictive power but the ultimate method to incorporate this information remains a work in progress. Novel methods to develop a personalized prediction model that provides outcomes that are specific for a patient are currently under way and may change how we think about risk stratification in MDS patients in the future.