14-Protein Blood Signature Predicts Lung Cancer Risk Years Before Diagnosis
Researchers from the Francis Crick Institute and University College London have discovered a blood protein signature consisting of 14 proteins that can forecast lung cancer risk over five years prior to diagnosis. This research, featured in Cell, utilized machine learning techniques on data from 48,000 participants in the UK Biobank. The identified signature indicates an inflammatory condition in the lungs associated with a newly recognized cell type known as KAC cells. In mouse studies, inhibiting the inflammatory molecule IL-1β led to a decrease in KAC cells and decelerated tumor growth. A reexamination of the CANTOS trial revealed that patients exhibiting a high signature who received canakinumab experienced almost a 50% reduction in lung cancer risk. This test may be accessible to high-risk individuals within two years. This research builds upon the TRACERx project, spearheaded by Prof Charles Swanton, who has recently been elected a Fellow of the Science Museum Group.
Key facts
- 14-protein blood signature predicts lung cancer risk >5 years before diagnosis
- Study published in Cell by Francis Crick Institute and UCL
- Machine learning applied to UK Biobank data from 48,000 volunteers
- Signature linked to KAC cells and inflammation (IL-1β)
- Blocking IL-1β in mice reduced KAC cells and slowed tumor growth
- CANTOS trial reanalysis: high-signature patients on canakinumab had ~50% lower lung cancer risk
- Test could be available in 2 years for high-risk patients
- Prof Charles Swanton led the study; elected Fellow of Science Museum Group
Entities
Institutions
- Francis Crick Institute
- University College London
- Cell
- UK Biobank
- Cancer Research UK
- Rosetrees Trust
- Mark Foundation for Cancer Research
- Science Museum Group
- CANTOS
Locations
- London
- United Kingdom