The Myelofibrosis AlloRisk Stratifier has been meticulously developed by leveraging the comprehensive EBMT database, which encompasses a wide array of myelofibrosis patients who have undergone allogeneic hematopoietic cell transplantation (allo-HCT) between 2005 and 2020. This robust foundation has been augmented with sophisticated machine learning algorithms, enabling the tool to accurately predict individual patient risk levels based on 10 variables taken at the time of transplant.
This innovative model serves as a powerful instrument for inferring patient risk, providing invaluable insights into the potential outcomes following allo-HCT. However, it's important to note that while the AlloRisk Stratifier is adept at risk assessment, it is not designed to dictate specific transplant conditioning protocols or graft-versus-host disease (GVHD) prophylaxis strategies. Its primary objective is to inform and enhance clinical decision-making by offering a clearer understanding of risk profiles.
Patients are classified in 4 different groups:
By incorporating this risk stratification tool into the clinical workflow, healthcare professionals can achieve a more nuanced understanding of patient risks, thereby facilitating more tailored and informed treatment planning. This approach ultimately aims to optimize patient care and improve the overall success rates of allo-HCT in the myelofibrosis patient population.
Patients with myelofibrosis (MF) undergoing allo-HCT who fall above the 93rd percentile indicated by the purple arrow, exhibit an overall survival of less than 50% at 1 year. Similarly, those above the 70th percentile indicated by the red arrow, demonstrate an overall survival of less than 50% at 2 years.
Age
HCT-CI Score
Karnofsky Performance Status
Leukocytes (10^9/L)
Blood Blasts (%)
Hemoglobin (g/dL)
Platelets (10^9/L)
Donor Type
Type of conditioning
Type of GVHD prophylaxis
The risk bar graphically represents a patient's risk level as a percentile, ranging from 0 to 100. A score of 0 signifies minimum risk, while 100 represents maximum risk. Visually, the bar evolves in color from green to red to indicate the level of risk: green denotes low risk, and red signifies high risk. This percentile and color-coded system provides an immediate, intuitive understanding of how a patient's risk compares to the broader population, assisting in informed medical decision-making.
The model predicts each patient's Non-Relapse Mortality (NRM) and overall Mortality Rate at 12 and 24 months. NRM represents the chance of death from non-relapse causes, while Mortality Rate covers all causes. For example, a 20% 24-month NRM means a 20% risk of dying from non-relapse causes in the first two years after transplantation.