Machine Learning Improves Risk Stratification in Myelodysplastic Neoplasms - An Analysis of the Spanish Group of Myelodysplastic Syndromes

Aug 5, 2026

Myelodysplastic neoplasms (MDS) are heterogeneous hematological stem cell disorders marked by dysplasia, cytopenias, and an elevated risk of acute leukemia, making accurate prognostication essential because outcomes and treatments range from observation to allogeneic stem cell transplantation. Registry data from 90 Spanish institutions yielded 7,202 patients split into training (80%) and test (20%) sets. Random survival forests modeled overall survival and leukemia-free survival using eight clinical, hematologic, and cytogenetic variables, outperforming IPSS-R and age-adjusted IPSS-R. Results were externally validated.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Abstract

Introduction

Disease overview and current classification

Need for accurate risk prognostication

Downloaded from [http://journals.lww.com/hemasphere by BhDMf5ePHKav1zEoum1tQfN4a](http://journals.lww.com/hemasphere by BhDMf5ePHKav1zEoum1tQfN4a)+kJLhEZgbsIHo4XMi0hCy wCX 1AWnYQp/ I l Qr HD3i 3 D0OdRyi 7TvS Fl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 11/03/2023 Article Open Access Machine Learning Improves Risk Stratification in Myelodysplastic Neoplasms: An Analysis of the Spanish Group of Myelodysplastic Syndromes Adrian Mosquera Orgueira1 , Manuel Mateo Perez Encinas1 , Nicolas A Diaz Varela2 , Elvira Mora3 , Marina Díaz-Beyá4 , María Julia Montoro5 , Helena Pomares6 , Fernando Ramos7 , Mar Tormo8 , Andres Jerez9 , Josep F Nomdedeu10 , Carlos De Miguel Sanchez11 , Arenillas Leonor12 , Paula Cárcel13 , Maria Teresa Cedena Romero14 , Blanca Xicoy15 , Eugenia Rivero16 , Rafael Andres del Orbe Barreto17 , Maria Diez-Campelo18 , Luis E. Benlloch19 , Davide Crucitti20 , David Valcárcel5 GRAPHICAL ABSTRACT Downloaded from [http://journals.lww.com/hemasphere by BhDMf5ePHKav1zEoum1tQfN4a](http://journals.lww.com/hemasphere by BhDMf5ePHKav1zEoum1tQfN4a)+kJLhEZgbsIHo4XMi0hCy wCX 1AWnYQp/ I l Qr HD3i 3 D0OdRyi 7TvS Fl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 11/03/2023 Article Open Access Machine Learning Improves Risk Stratification in Myelodysplastic Neoplasms: An Analysis of the Spanish Group of Myelodysplastic Syndromes Adrian Mosquera Orgueira1 , Manuel Mateo Perez Encinas1 , Nicolas A Diaz Varela2 , Elvira Mora3 , Marina Díaz-Beyá4 , María Julia Montoro5 , Helena Pomares6 , Fernando Ramos7 , Mar Tormo8 , Andres Jerez9 , Josep F Nomdedeu10 , Carlos De Miguel Sanchez11 , Arenillas Leonor12 , Paula Cárcel13 , Maria Teresa Cedena Romero14 , Blanca Xicoy15 , Eugenia Rivero16 , Rafael Andres del Orbe Barreto17 , Maria Diez-Campelo18 , Luis E. Benlloch19 , Davide Crucitti20 , David Valcárcel5 Correspondence: David Valcárcel ([dvalcarcel@vhio.net](dvalcarcel@vhio.net)); Adrián Mosquera Orgueira ([adrian.mosquera.orgeira@sergas.es](adrian.mosquera.orgeira@sergas.es)) . ABSTRACT Myelodysplastic neoplasms (MDS) are a heterogeneous group of hematological stem cell disorders characterized by dysplasia, cytopenias, and increased risk of acute leukemia. As prognosis differs widely between patients, and treatment options vary from observation to allogeneic stem cell transplantation, accurate and precise disease risk prognostication is critical for decision making. With this aim, we retrieved registry data from MDS patients from 90 Spanish institutions. A total of 7202 patients were included, which were divided into a training (80%) and a test (20%) set. A machine learning technique (random survival forests) was used to model overall survival (OS) and leukemia-free survival (LFS) . The optimal model was based on 8 variables (age, gender, hemoglobin, leukocyte count, platelet count, neutrophil percentage, bone marrow blast, and cytogenetic risk group) . This model achieved high accuracy in predicting OS (c-indexes; 0.759 and 0.776) and LFS (c-indexes; 0.812 and 0.845) . Importantly, the model was superior to the revised International Prognostic Scoring System (IPSS-R) and the age-adjusted IPSS-R. This difference persisted in different age ranges and in all evaluated disease subgroups. Finally, we validated our results in an external cohort, confirming the superiority of the Artificial Intelligence Prognostic Scoring System for MDS (AIPSS-MDS) over the IPSS-R, and achieving a similar performance as the molecular IPSS. In conclusion, the AIPSS-MDS score is a new prognostic model based exclusively on traditional clinical, hematological, and cytogenetic variables. AIPSSMDS has a high prognostic accuracy in predicting survival in MDS patients, outperforming other well-established risk-scoring systems. INTRODUCTION Myelodysplastic neoplasms (MDS) comprise diagnostic entities characterized by the presence a of variety of dysplasia, cytopenias and risk of progression to bone marrow (BM) failure or acute myeloid leukemia.1 Currently,

FAQs

Why is risk stratification critical in myelodysplastic neoplasms (MDS)?

MDS prognosis varies widely between patients, and treatment decisions depend on accurate risk prediction. Options can range from observation to allogeneic stem cell transplantation.

How was the machine learning model developed and tested?

Registry data from 90 Spanish institutions included 7,202 patients divided into an 80% training set and a 20% test set. A random survival forest model was used to predict overall survival and leukemia-free survival.

What variables did the optimal model use?

The best model was based on eight variables: age, gender, hemoglobin, leukocyte count, platelet count, neutrophil percentage, bone marrow blast, and cytogenetic risk group.

How did the proposed model compare with existing scoring systems?

The model showed higher accuracy than revised IPSS-R and age-adjusted IPSS-R across age ranges and disease subgroups. External validation confirmed superiority of AIPSS-MDS over IPSS-R and similar performance to molecular IPSS.

Aug 5, 2026

Upload by Olivia Brown