Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

In a study published in the Journal of Clinical Oncology, an international research team has used machine learning to improve risk stratification for patients over 60 diagnosed with acute myeloid leukaemia, an aggressive type of blood cancer.

A patient and a doctor consulting a tablet together in a clinic. © Yuri A/Shutterstock.com

An international collaboration led by Paresh Vyas (MRC Molecular Haematology Unit in the MRC Weatherall Institute of Molecular Medicine) and Peter Valk (Erasmus University MC Cancer Institute) has created a tool to help clinicians predict how older patients with acute myeloid leukaemia will respond to treatment. Their new online tool AML60+ will allow doctors to apply this in clinical practice easily.

Acute myeloid leukaemia (AML) is the most common aggressive blood cancer in adults. In the UK, around 3100 are diagnosed with it each year. AML is a diverse disease with very varied patient outcomes; clinicians use various indicators to predict who will benefit from treatment and how likely it is that the leukaemia will come back (this is called “risk stratification”). However, current predictive tools used by clinicians are based on younger patients, not those over the age of 60, for whom the disease is especially diverse. 

In a study published in the Journal of Clinical Oncology earlier this month, a team of researchers from the UK, Netherlands, Switzerland and Belgium created a risk stratification system for patients over the age of 60 to identify those who would benefit most from intensive treatment and stem cell transplantation.

The research team analysed data from almost 2000 older patients diagnosed with AML. They investigated the impact of various factors on survival, including age, gender, and genetic and chromosome changes found in the cancer cells. Using machine learning, the researchers identified 9 features that predict the outcome of treatment and used these features to separate patients into four risk groups: favourable, intermediate, poor and very poor. These groups were found to be better at predicting patient outcomes than existing tools, such as the commonly used “ELN 2022” risk stratification.

Speaking about the study, Professor Vyas said:

Acute Myeloid Leukaemia (AML) is the most common aggressive adult leukaemia. Like all cancers, AML is more common as we get older; and the incidence rises steeply at the age of 60. The most important curative treatment is intensive chemotherapy and often bone marrow transplantation. What was needed by healthcare professionals worldwide was an accurate estimate of how well patients over the age of 60 would fare with this type of treatment. 

In this publication, we now provide an analysis of how patients will fare depending on their clinical state and genetic characteristics of their leukaemia at diagnosis. We have made an easy-to-use online app that any patient and healthcare professional can use to help inform how best to treat patients. We included patients from the UK and many parts of Europe in the largest study of its kind.

Read the full paper here: https://doi.org/10.1200/JCO.23.02631