Researchers from the University of Geneva (UNIGE) have unveiled a groundbreaking method that could transform colorectal cancer screening. Utilizing machine learning, they established a comprehensive catalogue of human gut bacteria, precisely detailing how various microbial subgroups operate within the body. This innovative approach allows for the detection of colorectal cancer through simple stool samples, providing a non-invasive and cost-effective alternative to traditional methods. The findings, published in Cell Host & Microbe, also pave the way for a deeper understanding of how gut microbiota affects overall health and disease.
The Need for Improved Screening
Colorectal cancer is often diagnosed at advanced stages, limiting treatment options and underscoring the urgent demand for more accessible and less invasive screening techniques. This need is particularly pressing as incidences among younger adults continue to rise for reasons that are not yet fully understood.
While the role of gut microbiota in colorectal cancer has been established, translating this knowledge into practical medical applications has posed significant challenges. Variability among strains within the same bacterial species complicates the picture; some may promote cancer, while others remain harmless.
Targeting Microbiota Subspecies
Mirko Trajkovski, a full professor in the Department of Cell Physiology and Metabolism at UNIGE, led the research. He explains, "We shifted our focus to the subspecies level of microbiota, which captures critical functional differences among bacteria, while being broad enough to identify variations across diverse populations."
Leveraging Machine Learning for Insights
The study involved analyzing vast amounts of biological data. Matija Trickovic, a PhD student and the study's first author, noted, "We successfully created the first extensive catalogue of human gut microbiota subspecies and an efficient method to utilize it in both research and clinical settings."
A Stool Test That Competes with Colonoscopy
By integrating their bacterial catalogue with existing clinical datasets, the researchers developed a model capable of identifying colorectal cancer solely through stool samples. The results were impressive, with the method detecting 90% of cancer cases--closely rivaling the 94% detection rate of colonoscopies and outperforming all current non-invasive alternatives.
With further clinical data, this model could achieve even greater accuracy, potentially matching colonoscopy's effectiveness. Such a test could streamline routine screenings, reserving colonoscopies for confirming positive results.
Broader Implications for Health Diagnostics
A clinical trial is in the works in collaboration with Geneva University Hospitals (HUG) to explore which cancer stages and lesions can be detected using this method. The potential applications extend well beyond colorectal cancer; by studying subspecies differences, researchers aim to uncover how gut microbes influence various health conditions.
"Our method may soon facilitate the development of non-invasive diagnostic tools for a wide array of diseases, all stemming from a single microbiota analysis," concludes Trajkovski.