Acute myeloid leukemia (AML) is an aggressive blood cancer with a high risk of early death and relapse, even with today’s best treatments. Clinicians currently rely on risk assessment tools like the European LeukemiaNet (ELN) system to guide care, but these models are based primarily on information available at diagnosis and fail to capture how the disease evolves over time. One critical, yet often overlooked, factor is inflammation, which is a natural immune response that can influence how aggressively the disease progresses and how well a patient responds to treatment. In this study, our team will develop a new predictive system called FIRE-AML (A Longitudinal, Machine Learning-Based Frame...
Read More
Acute myeloid leukemia (AML) is an aggressive blood cancer with a high risk of early death and relapse, even with today’s best treatments. Clinicians currently rely on risk assessment tools like the European LeukemiaNet (ELN) system to guide care, but these models are based primarily on information available at diagnosis and fail to capture how the disease evolves over time. One critical, yet often overlooked, factor is inflammation, which is a natural immune response that can influence how aggressively the disease progresses and how well a patient responds to treatment. In this study, our team will develop a new predictive system called FIRE-AML (A Longitudinal, Machine Learning-Based Framework of Inflammation Driving Relapse and Early Mortality in AML). This tool, which will be made publicly available and freely accessible, will integrate inflammatory signals in the blood with genetic, clinical, and treatment data collected at multiple points over the course of a patient’s treatment. Using advanced but practical machine learning models, FIRE-AML will help identify patients at highest risk of early death or relapse. By enabling earlier and more accurate risk assessment, this approach will empower clinicians to tailor treatment strategies more precisely and intervene sooner to improve outcomes. We will study three key aims: 1 -To better predict early death from AML using blood inflammation markers alongside other patient data. 2- To monitor changes in inflammation and other data over time to catch signs of relapse or treatment resistance earlier, and 3- To understand how inflammation contributes to treatment failure by studying patient samples in lab-grown bone marrow mouse models known as 'organoids'. If successful, FIRE-AML could change the way we care for AML patients. It may allow clinicians to personalize treatments more effectively, minimize toxicity and improve survival.
Read Less
|
