A Laval University researcher and colleagues at McGill University have developed a method that uses artificial intelligence to determine with 95% accuracy the risk of lung cancer recurrence in a patient who has already had a lung tumor removed.
This approach was developed by pathologist and professor at the University of Laval Faculty of Medicine Philippe Joubert and his colleagues Logan Walsh and Daniela Quail from McGill University. To achieve this, they analyzed tumors from more than 400 patients who had undergone lung resection – that is, surgical removal – from the biobank of the University Institute of Cardiology and Pulmonology of Quebec (IUCPQ).
“We wanted to exploit the composition of what we call the tumor microenvironment,” explains Dr. Joubert. We profiled using technology that allowed multiple markings to be made simultaneously on the same blade.”
Dr. Philippe Joubert, pathologist and professor at the Faculty of Medicine at Laval University. Decency
The use of artificial intelligence has made it possible to process data and extract predictors and markers that predict cancer recurrence in people who have already undergone surgery.
This valuable information allows researchers to predict with approximately 95% accuracy whether a person who has had a lung tumor removed will have a recurrence of cancer using a tumor sample of just under 1 mm2.
A promising breakthrough
This is a significant advance considering that approximately 40% of people who have had surgery for lung cancer experience a recurrence. This new approach will make it possible to quickly identify them after surgery and offer them the most appropriate treatments for their situation. In addition to improving forecasts, this also enables better use of resources in healthcare facilities and enables more targeted targeting of patients at risk.
“Tests with this level of performance, I can’t remember seeing many,” enthuses Dr. Joubert. I was lying on the floor when I saw the results. It’s very promising.”
However, we will have to wait for the use of this approach in healthcare facilities. The test needs to be simplified first and perhaps tested prospectively in real populations to see if the performance can be replicated retrospectively.
Daniela Quail, Professor at the Rosalind and Morris Goodman Cancer Institute at McGill University. With kind approval
Logan Walsh, Professor at the Rosalind and Morris Goodman Cancer Institute at McGill University. With kind approval
This breakthrough is one of the discoveries of the annual magazine Québec Science.
Can you share information about this story?
Write to us or call us directly at 1 800-63SCOOP.