Scientists have been working with pathologists to build an artificial intelligence tool that can predict which patients with lung cancer have a higher risk of the disease coming back after treatment. The tool was able to differentiate between cancer and immune cells, which enabled the researchers to better understand how lung cancers evolve in response to the immune system in individual patients.
Although its utility research is still in the early stages, the approach could speed up how doctors predict which patients are more likely to have their cancer return and can help them tailor their treatment plans.
A tool to quantify immune “hot and cold regions”
The AI tool was developed by researchers at The Institute of Cancer Research in collaboration with scientists at the University College London Cancer Institute and the Francis Crick Institute.
The tool was trained by pathologists to pick out immune cells from the cancer cells. This maps out areas in tumours where the number of immune cells was higher compared to cancer cells, which allowed the team to identify tumour areas packed with immune cells, “hot regions”, and other parts which were devoid of them – “cold regions”.
Following this, the researchers found that patients with a higher number of cold regions had a higher risk of relapse. The paper was published in Nature yesterday and is part of the TRACERx (Tracking Cancer Evolution through therapy Rx) lung study, funded by Cancer Research UK.
The team, led by Dr YinYin Yuan, combined the AI pathology image-mapping tool with next generation sequencing (NGS) and compared the extent to which the patient’s tumour genetic makeup differed when comparing immune hot or cold areas within the same tumours, giving clues to how they had changed in response to evolutionary pressures.
The work revealed that cancer cells in the immune “cold” regions may have evolved more recently than those in the immune hot regions. The researchers suggested that the cold areas may have developed a cloaking mechanism under evolutionary pressure to hide from the body’s natural immune defences.
The tool can assess how many regions in the tumour that are being “hidden” from the immune system which is crucial as these are associated with relapse. Dr Yinyin Yuan, one of the authors and team leaders at ICR said that they “applied artificial intelligence with genetic data and pathology images to create the tool, that could in the future help pick out patients who are at highest risk of their cancer coming back” and that their research has given fresh insights into why some lung cancers are difficult to treat.
