Published on
Updated
Reading 2 mins.
Artificial intelligence would predict the risks of spreading aggressive breast cancer, based on changes in patients’ lymph nodes. An asset for the management of triple negative breast cancers.
Published today in The Journal of Pathology, the English study shows that by analyzing immune responses in the lymph nodes of women with triple-negative breast cancer, it is possible to determine the likelihood of the disease spreading to other parts of the body.
Triple negative breast cancer: a dreadful cancer
About 15% of breast cancers are said to be “triple negative” (because they do not respond to hormonal treatments or anti-HER2 treatments). Faced with these formidable tumours, which most often strike young women, there are currently few targeted treatments. Triple-negative breast cancer is more likely than most other breast cancers to come back or spread to other parts of the body in the first few years after treatment.
In this case, it is called metastatic breast cancer and although it is treatable, it cannot be cured. King’s team developed an AI model to predict a patient’s likelihood of developing metastatic (incurable) breast cancer based on immune responses in the lymph nodes.
Study the lymph nodes, guardian of our defense system
Lymph nodes are the size of a pea and are found throughout the body as they help the body fight infections. Breast cancer cells usually spread first to the lymph nodes in the armpit (armpit) which are closest to the tumor. If this happens, patients usually receive more aggressive treatment.
However, the scientists found that even when the breast cancer cells had not spread to the lymph nodes, it was still possible to predict from their immune responses the likelihood of the cancer spreading elsewhere in the body.
The scientists tested their AI model on more than 5,000 lymph node sections stored in biobanks and donated by 345 patients. They then used the technique of deep learning or “deep learning” which is a sub-branch of artificial intelligence. It aims to build knowledge from a large amount of information (the greater the amount of information, the more knowledge will be refined). In the case that interests us, it is a question of “feeding” an algorithm with numerous images of lymph nodes and of having in parallel the medical history of the patients.
Result: their model confirmed that it could establish the probability of spread of breast cancer to other organs.
Lead author of the study, Dr Anita Grigoriadis concludes: “By demonstrating that lymph node changes can predict whether triple-negative breast cancer will spread, we built on our growing knowledge of the important role the immune response can play in understanding a patient’s prognosis. We plan to further test the model in other European centers to make it even more robust and accurate. The transition from assessing tissue on glass slides under a microscope to using computers in the healthcare system (NHS) is accelerating. We want to take advantage of this change to develop AI software based on our model that pathologists can use to benefit women with this difficult-to-treat breast cancer.“.