Supplementary Material for: IPSS, IPSS-R and IPSS-M had no identical prognostic power depending on treatment in myelodysplastic syndromes

Introduction The Revised International Prognostic Scoring System (IPSS-R) has long been the standard prognostic tool in the management of myelodysplastic syndromes (MDS). Recently, a new clinical-molecular prognostic model, the IPSS-M, was introduced. This model integrates MDS-related gene mutations and offers improved accuracy and precision in predicting patient outcomes and survival. In this study, we compare the prognostic value and predictive capacity of these prognostic scores across different treatment groups. Methods IPSS, IPSS-R, and IPSS-M scores were calculated at diagnosis for all eligible MDS patients in this cohort. Sankey diagrams were generated to visually compare the risk stratification across the three scoring systems. Results A total of 394 patients were included in this retrospective analysis. During the course of the disease, 59.1% required treatment, with 38.1% (150 patients) receiving erythropoiesis-stimulating agents (ESA) and 19.0% (75 patients) treated with hypomethylating agents (HMA). The IPSS-M was calculated for 281 patients and reclassified 41.9% of cases: 20.3% (n=57) were upstaged and 21.7% (n=61) were downstaged. All three scoring systems demonstrated significant stratification of overall survival (OS) in the global cohort (p<0.001 for IPSS, IPSS-R, and IPSS-M). Notably, IPSS failed to predict OS in patients treated with azacitidine, and none of the scores successfully predicted OS in patients treated with ESA. Conclusion The IPSS-M effectively reclassified 41.9% of patients, upstaging 20.2% and downstaging 21.7%. However, our analysis found that none of the prognostic scores effectively predicted outcomes for ESA-treated patients. Further studies are needed to assess whether the predictive value of these scoring systems is influenced by the specific treatment modalities used.