Researchers have identified a 14-protein blood signature capable of predicting lung cancer risk more than five years before diagnosis, potentially opening the door to a new era of precision cancer prevention.
The study, published in Cell by investigators at The Francis Crick Institute and University College London (UCL), combines large-scale human population data, mechanistic laboratory studies, and clinical trial analyses to demonstrate that a blood-based inflammatory signature can identify individuals at elevated risk of lung cancer and may pinpoint those most likely to benefit from preventive treatment.
The findings move beyond traditional risk models based on age and smoking history and offer what researchers describe as a potential equivalent of cholesterol testing for lung cancer prevention.
Moving beyond smoking-based risk assessment
Current lung cancer screening programs primarily target older individuals with a history of smoking. While smoking remains the leading risk factor, many lung cancers arise in people who would not qualify for screening under existing criteria, including never-smokers and individuals exposed to environmental pollutants.
The research team sought to develop a biologically informed method of identifying risk by focusing on inflammation, which has emerged as a critical driver of tumor development.
Previous work from the group demonstrated that air pollution can promote lung cancer by triggering inflammatory responses that awaken dormant cells carrying cancer-causing mutations. The new study aimed to determine whether this inflammatory state could be detected in the blood before cancer becomes clinically apparent.
Machine learning identifies a 14-protein signature
Using plasma protein measurements from more than 48,000 participants in the UK Biobank, researchers applied machine learning approaches to identify blood proteins associated with future lung cancer diagnoses.
The algorithm incorporated conventional risk factors such as age, smoking status, prior lung disease, and plasma protein profiles. Analysis revealed a panel of 14 circulating proteins that consistently predicted lung cancer risk within five years of diagnosis.
The signature was validated across eight independent international datasets and remained predictive across diverse populations, including a cohort composed entirely of non-smokers.
Individuals who later developed lung cancer consistently exhibited elevated levels of the signature years before their diagnosis.
“This is a proof of concept that, one day, we could use this signature to offer preventive treatment to people at risk of lung cancer,” said Tej Pandya, clinical PhD student at UCL and visiting scientist at The Francis Crick Institute.
Detecting an inflammatory state before cancer emerges
One of the study’s most significant findings is that the signature appears to reflect a pre-cancerous inflammatory environment rather than the presence of an undetected tumor.
The researchers found evidence that the protein profile originates from changes within the lung microenvironment before malignant transformation occurs.
Further analyses showed that the same signature was elevated in individuals who later developed chronic obstructive pulmonary disease (COPD) or idiopathic pulmonary fibrosis, suggesting it may identify a broader inflammatory state that predisposes individuals to multiple lung diseases.
Studies in mouse models provided additional support for this hypothesis. Exposure to air pollution increased both the protein signature and the abundance of a cellular state known as KAC cells—adaptive cells that emerge during tissue injury but can become malignant when cancer-driving mutations are present.
Mutant cells arising from several distinct lung cell populations converged on this same KAC state during the earliest stages of cancer development.
Linking air pollution, inflammation, and cancer
The findings build on earlier research implicating the inflammatory cytokine interleukin-1 beta (IL-1β) as a critical mediator of pollution-driven lung cancer.
The investigators demonstrated that air pollution exposure increased IL-1β signaling, elevated components of the 14-protein signature, and expanded KAC cell populations.
Blocking IL-1β in mice reduced KAC cell numbers and slowed early tumor formation, providing experimental evidence that inflammatory signaling contributes directly to cancer initiation.
These observations suggest that the blood signature may serve not only as a risk marker but also as a biological indicator of an underlying process that can be therapeutically targeted.
Revisiting a major clinical trial
To determine whether the signature could identify patients most likely to benefit from preventive intervention, the researchers revisited data from the landmark CANTOS trial.
The trial originally evaluated the IL-1β inhibitor canakinumab for cardiovascular disease prevention and unexpectedly reported reduced lung cancer incidence as an exploratory outcome. However, the overall cancer-prevention benefit appeared too modest to justify widespread use of the drug for this purpose.
The new analysis tells a different story.
Researchers examined data from 4,651 CANTOS participants and found that individuals with elevated levels of the 14-protein signature experienced the greatest benefit from canakinumab treatment. In this high-risk subgroup, lung cancer incidence was nearly cut in half.
By restricting treatment to those identified by the biomarker signature, the number needed to treat to prevent one lung cancer case fell to 55, a figure comparable to widely accepted cardiovascular prevention strategies such as statin therapy.
Toward precision cancer prevention
The work represents a shift in how researchers think about cancer prevention.
Rather than treating large populations indiscriminately, the study suggests that molecular biomarkers could identify individuals in a reversible pre-disease state and guide targeted interventions before cancer develops.
“Drugs like statins have transformed the prevention of cardiovascular disease, used to treat individuals with a high low-density lipoprotein (LDL),” said Charlie Swanton, FRCP, PhD, clinical research director at The Francis Crick Institute and professor of cancer at UCL.
“But we don’t yet have an LDL-like marker of risk or a statin for lung cancer.”
Swanton added that identifying an inflammatory state before tumor formation provides a potential “window of opportunity” in which preventive treatment could be most effective.
Implications beyond lung cancer
The investigators note that the inflammatory signature may reflect a broader biological phenomenon associated with aging and chronic disease.
Because the signature was also associated with future COPD and pulmonary fibrosis, it may represent a shared pre-symptomatic inflammatory state that precedes multiple age-related lung disorders.
If validated in prospective studies, the approach could ultimately support routine blood-based risk assessment and targeted prevention strategies not only for lung cancer but potentially for other inflammation-driven diseases.
For now, the findings provide one of the strongest demonstrations yet that cancer risk can be detected years before diagnosis and that those biological signals may be actionable.
The challenge ahead will be determining whether identifying high-risk individuals and intervening early can translate into measurable reductions in lung cancer incidence, a question future prospective prevention trials will seek to answer.
The post Blood-Based Risk Score Could Enable Earlier Lung Cancer Prevention appeared first on Inside Precision Medicine.
imaged capillary isoelectric focusing (icIEF) fractionation system with Refeyn’s TwoMP mass photometry platform. The combined solution makes it possible for researchers to directly correlate charge heterogeneity with molecular weight and aggregation at single‑molecule resolution in four hours.