Colorectal cancer is the second leading cause of cancer-related deaths worldwide.
When found early, it can be treated effectively, but the main screening method—colonoscopy—can be uncomfortable, invasive, and costly.
As a result, many people delay testing, and the disease is often discovered too late. Now, researchers at the University of Geneva (UNIGE) have found a promising alternative: using gut bacteria in stool samples as a non-invasive, low-cost screening tool.
Published in Cell Host & Microbe, the study shows how machine learning can identify colorectal cancer by analyzing the bacteria living in the human gut.
For the first time, scientists created a detailed map of gut bacteria down to the “subspecies” level. This level of detail is important because different strains of the same species can act in very different ways—some may contribute to cancer development while others may not.
“Looking only at bacterial species misses important differences, while looking at every single strain is too specific and varies too much from person to person,” explained Professor Mirko Trajkovski from UNIGE’s Faculty of Medicine.
“By focusing on subspecies, we capture the meaningful differences that affect health while still being able to compare across groups and populations.”
To build this bacterial catalog, researchers analyzed massive amounts of data using advanced bioinformatics and machine learning.
The result was the first complete inventory of human gut microbiota subspecies, along with a method for using it in both research and clinical practice. They then combined this catalog with clinical data to train a predictive model that can detect colorectal cancer from stool samples.
The results were impressive. The new method correctly identified 90% of cancer cases. That’s nearly as good as colonoscopy, which detects about 94%, and much better than other non-invasive screening tools currently available.
“We were confident in our approach, but the results were still striking,” said Matija Trickovic, the study’s first author.
If more clinical data are added, the model could become even more accurate and potentially match colonoscopy. In the future, this could mean that patients only undergo colonoscopies when stool-based testing indicates a high likelihood of cancer, saving time, cost, and discomfort for millions.
The researchers are already preparing a clinical trial with Geneva University Hospitals to better determine which stages of cancer and types of lesions can be detected. But the potential applications go even further. Since the method can capture how bacterial subspecies function in the body, it may help researchers understand the role of the microbiome in many other diseases, not just cancer.
“This approach opens the door to developing simple, non-invasive diagnostic tools for a wide range of illnesses, all based on a single stool test,” said Trajkovski.
If successful, gut bacteria may soon become an everyday tool for protecting health and saving lives.
If you care about cancer risk, please read studies that exercise may stop cancer in its tracks, and vitamin D can cut cancer death risk.
For more health information, please see recent studies that yogurt and high-fiber diet may cut lung cancer risk, and results showing that new cancer treatment may reawaken the immune system.
Source: KSR.
