Most current cancer tests work by detecting biomarkers—molecules in the body that indicate a disease is present.
These biomarkers are often produced by the tumour itself or related proteins. The problem is that such markers tend to be more abundant only after the cancer has grown significantly. By then, treatment options may be more limited, and the disease harder to control.
Now, scientists at the Spanish National Cancer Research Centre (CNIO), led by Professor Gonçalo Bernardes, have developed a new blood test that can detect solid tumours at a much earlier stage.
Not only that—it can also give doctors valuable information about which treatments might work best for each patient. Their study has been published in Nature Communications.
Instead of looking for substances made by the tumour, the research team focused on the body’s immune response to cancer. Since the 19th century, doctors have known that the appearance of cancer cells triggers changes in the immune system.
These changes are strongest in the early stages of cancer, but until now, they had never been used in a diagnostic test. Bernardes and his team studied changes in blood proteins that happen when the immune system is disrupted by cancer.
“Our approach has proven particularly effective in detecting tumours early,” says Bernardes. “If we catch them early, we can treat many types of cancer more successfully.”
One challenge was that human blood contains over 5,000 different proteins, making it extremely complex to analyse. To simplify the task, the researchers used bioinformatics to narrow their focus to five specific amino acids: lysine, tryptophan, tyrosine, cysteine, and cysteine not bound to disulphide bonds.
They used a chemical process called fluorogenic reactions, where the amino acids emit light when exposed to certain conditions, allowing scientists to measure their exact concentrations. With the help of artificial intelligence—specifically, machine learning—the team analysed these measurements and found patterns that could indicate the presence of cancer.
The test was tried on samples from 170 patients. It correctly detected 78% of cancers and, importantly, had a 0% false positive rate—meaning it did not wrongly identify healthy people as having cancer.
Bernardes explains that the test is easy to use and only needs a small blood sample. The reagents required are simple and available in most hospitals. The team is also developing a software platform to analyse the results, making it practical for widespread use.
The study also found that the test could distinguish between signals from cancer and signals from other diseases, like COVID-19. It could even tell apart different types of cancer and their stages. This opens the door for the test to be used not just for diagnosis but also for treatment planning.
In one example, the test predicted with 100% accuracy that a patient would not respond to a certain anti-metastatic treatment, and it was 87% accurate in predicting positive responses. This could make it a valuable tool for precision medicine—choosing the right treatment for the right patient.
While the results are promising, the researchers note that larger studies are needed before the test can be widely used. Two clinical trials are already taking place in the UK, funded by the National Health Service, and more trials are underway in the US and China.
Once fully developed, the test will be commercialised by Proteotype Ltd, a company co-founded by Bernardes and his collaborators.
This new approach represents a shift from traditional cancer diagnostics, which focus on tumour-produced biomarkers. By targeting the immune system’s early response, the test may be able to detect cancers well before current methods can.
The fact that it also predicts treatment response adds major value for personalised medicine. The 0% false positive rate is especially important, as false alarms can cause unnecessary stress and costly follow-up procedures.
However, the sample size of 170 patients is small, and larger, diverse studies will be needed to confirm these results. If the current trials succeed, this could become a game-changing tool for early cancer detection and treatment planning worldwide.
Copyright © 2025 Knowridge Science Report. All rights reserved.
