Researchers at UCL and the Francis Crick Institute have identified 14 proteins in the blood that can predict lung cancer risk more than five years before diagnosis. The findings, published in Cell, mean that people at higher risk of developing lung cancer who might benefit from drugs to prevent the disease taking hold could be identified early – a major step towards precision cancer prevention.
The study involved analysis of blood samples from more than 48,000 people in the UK Biobank using machine learning. Along with age, smoking status and previous history of lung disease, the team identified 14 key proteins in the blood that could predict a future diagnosis of lung cancer. The team validated these proteins in eight datasets from across the world, finding that levels of the protein signature were higher in people who later developed lung cancer.
Analysis of this protein signature suggested that it does not come from the tumour itself but reflects an altered inflammatory lung environment that precedes cancer. This previous inflammation could come about because of factors such as air pollution or cigarette smoke.
Earlier work from this research team demonstrated for the first time how air pollution can cause lung cancer in people who have never smoked. The latest study found that pollution can increase both the protein signature and the number of abnormal lung cells linked to early cancer development.
The study, funded by Cancer Research UK and supported by the National Institute for Health and Care Research UCLH Biomedical Research Centre, suggests that drugs which block the action of an inflammatory molecule called IL-1β could help prevent lung cancer in people with the high-risk protein signature.
The work is part of the landmark TRAcking Cancer Evolution through therapy (Rx) (TRACERx) lung cancer studies, led by Professor Charles Swanton, Chair in Personalised Cancer Medicine at theUCL Cancer Institute.
Professor Swanton, who is also Clinical Research Director and Principal Group Leader at the Crick, said: “Drugs like statins have transformed the prevention of cardiovascular disease, used to treat individuals with a high ‘low density lipoprotein’ (LDL). But we don’t yet have an LDL-like marker of risk or a statin for lung cancer. In the clinic, we see first-hand the impact of diagnosing lung cancer at a late stage, so being able to identify people at greater risk and intervene before the disease develops is critical.
“Finding a signal for an inflammatory state in the lungs has given us insight into this window of opportunity, when preventative treatment could work best. This work supports a relatively new idea in the field, that some common age-related diseases, causing a high burden of disease in the community, share a common, presymptomatic state of inflammation. We think the signature could in the future help to predict and help prevent lung cancer and other lung diseases.”
Tej Pandya, Clinical PhD Student at UCL Institute of Health Informatics and visiting scientist at the Crick , said: “We used machine learning on plasma data from over 48,000 people to identify the 14-protein signature, and it has been incredible to validate it across eight datasets with more than 80 collaborators on five continents.
“Working hand-in-hand with scientists in the lab to understand the biology in mouse models, we've shown that the signature reflects an altered inflammatory lung environment before cancer takes hold. It's a proof of concept that, one day, we could use this signature to offer preventive treatment to people at risk of lung cancer.”
Hayley Brown, Research Information Manager at Cancer Research UK said: "By revealing the earliest warning signs of cancer, this research brings us closer to intervening sooner and potentially stopping the disease before it starts.
“In doing so, it could help spare people and their loved ones from the impact of a cancer diagnosis, treatment, and everything that follows, allowing them instead to focus on the moments that matter most."
- View the paper in Cell
