Real-life challenges using personalized prognostic scoring systems in acute myeloid leukemia

Area of Science:

• Hematology
• Oncology
• Genetics

Background:

• Personalized medicine presents challenges in acute myeloid leukemia (AML) treatment.
• Genetic mutations identified in AML trials led to the development of the Knowledge Bank (KB) prognostic scoring algorithm.

Purpose of the Study:

• To evaluate the prognostic accuracy of the KB algorithm in a real-world cohort of 167 AML patients.
• To compare KB-predicted outcomes with actual patient outcomes.

Main Methods:

• Retrospective analysis of 167 AML patient data.
• Comparison of KB algorithm predictions against observed overall survival (OS).

Main Results:

• For AML patients under 60, OS aligned with favorable and intermediate European LeukemiaNet (ELN) risk categories.
• The KB algorithm inaccurately predicted OS for younger patients in the adverse ELN risk category.
• The KB algorithm also failed to predict OS for older patients (over 60) in the favorable ELN risk category.

Conclusions:

• Discrepancies in KB predictions may stem from new therapeutic options and improved allogeneic stem cell transplantation (aHSCT) outcomes.
• Prognostic models are crucial for personalized medicine in AML.
• Prospective validation of scoring systems is essential to incorporate recent therapeutic advancements in AML.