Early cancer clues found in multiple blood protein changes
Advances in monitoring could flag risks of diseases years before diagnosis
Scientists have built a potential early warning system to flag the risk of diseases such as cancers by monitoring the levels of thousands of blood proteins.
An international team of more than 100 researchers found an apparent biological marker for lung cancer years before diagnosis, a striking result they said now needed to be tested to see if it would be clinically useful.
The central idea of the project, known as the Human Disease Blood Atlas, is to gather comprehensive data to detect patterns of multiple protein changes that indicate specific diseases. The researchers argue this should help overcome the problem that variations in levels of an individual protein can indicate many different medical conditions, limiting their usefulness in diagnosis.
“We can separate universal ‘alarm bells’ . . . from truly disease specific signals, which is a crucial step for building blood tests that work in the clinic,” said Professor Mathias Uhlén, senior author of a paper on the work published in the journal Science on Thursday.
The project shows advances in blood analysis methods since the scandal of testing company Theranos during the 2010s. In 2022, Theranos founder Elizabeth Holmes was sentenced to more than 11 years in prison for defrauding investors, after falsely claiming to be able to perform many sophisticated diagnostic tests using a single drop of blood.
Using artificial intelligence models, the team scanned for 59 diseases by monitoring levels of up to 5,400 proteins in more than 8,000 people, according to the paper. It draws on the Human Protein Atlas, a Sweden-based programme directed by Uhlén that was launched in 2003 to map all human proteins in cells, tissues and organs using various state-of-the-art technologies.
Each adult has a distinctive stable blood protein baseline that is a molecular fingerprint for wellness and can be tracked for deviations, the researchers said.
The scientists found some protein profiles showed much raised levels in individuals found to have breast, ovarian, prostate, bowel and lung cancers. For lung cancer, the elevated readings could be seen several years before diagnosis.
The human blood atlas research was a “significant leap forward in our ability to predict disease risks using blood proteins”, said Lei Lu, lecturer in health data science and artificial intelligence at King’s College London, who was not involved in the research.
“This powerful approach allows us to see a comprehensive picture for the first time, distinguishing between protein patterns that are truly specific to a condition, such as a certain type of cancer, and those that are shared across many diseases,” Lu said. “This dramatically improves the potential accuracy and reliability of future risk prediction models.”
The atlas reflects progress in proteomics — the study of the proteins that are the building blocks of life — enhanced by artificial intelligence models. The UK Biobank genetic database and 14 drug companies this year launched a big proteomics project to identify disease subtypes more precisely, allowing treatments to be tailored and timed for maximum effectiveness.
Proteomics is allowing scientists to move beyond existing methods of analysing genes to examining the proteins whose production the genes instruct. This gives a more sophisticated and dynamic view of how the human body works and can malfunction, experts say.
“Unlike genomes which are static, proteomes are constantly changing so can help to flag changes within the body due to ill health,” said Hayley Wilson, a biomedical science policy analyst at the PHG Foundation not-for-profit think-tank. “The atlas will help us to understand how and when to intervene at an early stage of disease — and who could most benefit for a given disease.”