Healthcare (Commonwealth Union) – Taking into account that colorectal cancer ranks as the second most common cause of cancer-related deaths across the globe, it has been a key focus among researchers across the world. When caught early, it can often be treated successfully, but the high cost and discomfort of colonoscopies—the main diagnostic method currently used—frequently lead to delayed detection.
Scientists from the University of Geneva, have at present, made use of machine learning to map, the human gut microbiome in high detail, which is a first. This paves the way for a deeper understanding of the biological significance of various microbial subgroups. This comprehensive catalogue was then applied to mark colorectal cancer via bacteria found in simple stool samples, which makes available a non-invasive and affordable screening alternative. The implications are broad, this is beneficial not just for the detection of other cancers but for the advancing knowledge of the relationship between gut microbes and overall health as well. The study appears in Cell Host & Microbe.
Because colorectal cancer is often discovered only in its later stages, when treatment options are limited, there is an urgent need for easier, less invasive screening techniques—especially given the still poorly understood rise in cases among young adults. Although the role of gut microbiota in colorectal cancer has been recognized for years, turning that insight into practical medical applications has been difficult. One reason is that even within a single bacterial species, strains may behave very differently: some encourage cancer development, while others appear harmless.
Mirko Trajkovski, professor in the Department of Cell Physiology and Metabolism and the Diabetes Centre at the University of Geneva, Faculty of Medicine, who led the project pointed out that to overcome this challenge, they shifted focus away from simply analyzing bacterial species, which misses important variations, or strains, which differ widely between individuals, instead, they looked at the intermediate level—the subspecies. He further indicated that this approach provides the specificity needed to reveal how bacteria influence disease while still being broad enough to detect patterns across individuals, populations, and countries.
Matija Trickovic, a PhD student in Mirko Trajkovski’s lab and the lead author of the study indicated that the initial task was to process vast amounts of information. He indicated that for him as a bioinformatician, the main difficulty was to design a novel strategy for handling such large-scale data. Trickovic pointed out that they managed to build the first detailed catalogue of human gut microbiota subspecies, along with a reliable and accurate method that can be applied in both research and medical practice.
By integrating this catalogue with existing patient records, the researchers created a model capable of detecting colorectal cancer using only the bacterial composition found in stool samples.
“Although we were confident in our strategy, the results were striking,” said Matija Trickovic. “Our method detected 90% of cancer cases, a result very close to the 94% detection rate achieved by colonoscopies and better than all current non-invasive detection methods.”
By incorporating additional clinical data, this model could achieve even greater accuracy, approaching the reliability of colonoscopy. It holds the potential to become a standard screening method, enabling earlier detection of colorectal cancer—followed by colonoscopy only for a more targeted group of patients.
The first clinical trial, organized in partnership with Geneva University Hospitals (HUG), aims to more precisely determine which cancer stages and lesions can be identified. Yet the scope extends beyond colorectal cancer. By examining the variations between subspecies within the same bacterial species, scientists are now able to uncover how the gut microbiota affects human health.
Mirko Trajkovski indicated that this approach could soon pave the way for non-invasive diagnostic tests for many different diseases, all derived from a single microbiota analysis.
