Breast cancer breakthrough with new AI that can predict if disease will spread – and experts think it could save thousands of lives
- The technology measures the immune response of lymph nodes
- King’s College London researchers tested on 5,000, donated by 345 patients
Scientists have developed artificial intelligence which can predict if a patient’s breast cancer will spread.
The technology measures the immune response of lymph nodes, pea-sized lumps of tissue which help the body to fight infection.
Tests on the lymph nodes of people with triple negative breast cancer – an aggressive disease which is one of the most likely to spread or return – found it could predict if it was likely to metastasize.
Experts said the breakthrough could lead to more tailored treatment based on a woman’s individual risk profile, stopping the disease before it becomes incurable.
Researchers from King’s College London developed an AI model which they tested on more than 5,000 lymph nodes, donated by 345 patients.
Scientists have developed artificial intelligence which can predict if a patient’s breast cancer will spread (stock image)
Breast cancer cells typically first spread to lymph nodes in the armpit, or axilla, which are closest to the tumour.
When this happens, patients are usually given more intensive treatment to try and stop it developing elsewhere.
But scientists discovered that even when the breast cancer cells had not spread to the lymph nodes, their immune responses made it possible to predict the likelihood of the cancer spreading elsewhere.
They used a computer program to carry out image analysis of lymph nodes in cancer patients, which were then cross-referenced against the patient records and whether their breast cancer had spread.
Dr Anita Grigoriadis, who led the research at the Breast Cancer Now Unit at King’s College London, said: ‘Using our AI, we looked at lots of lymph node pictures and we homed in on specific patterns.
‘What we have seen is that when we look at many lymph nodes from many patients, we saw that when we find these features, they seem to be a sign that the patient somehow has the ability to hold off developing cancers in other organs for longer than in these patients which we have not found these features in the lymph nodes.’
She added: ‘By demonstrating that lymph node changes can predict if triple negative breast cancer will spread, we’ve built on our growing knowledge of the important role that immune response can play in understanding a patient’s prognosis.’
About 15 per cent of breast cancers are triple negative and there are currently few targeted treatments.
It is more common in women who have inherited an altered BRCA gene – made famous by Angelina Jolie – as well as black women, pre-menopausal and women aged under 40.
Publishing their findings in the Journal of Pathology, they hope to test the AI model in clinical trials.
Dr Simon Vincent, director of research, support and influencing at Breast Cancer Now, said: ‘Each year around 8,000 UK women are diagnosed with triple negative breast cancer, which is a more aggressive form of breast cancer, often with poorer outcomes.
‘If, thanks to this research, it’s possible to provide women with more tailored treatment and care based on the likelihood of the breast cancer spreading, it could help to save lives and reduce stress and worry. We look forward to further findings to understand how this could work in practice to benefit women affected by this type of breast cancer.’