New AI tool predicts cancer spread with surprising accuracy

Researchers at the University of Geneva have developed an artificial intelligence tool, MangroveGS, capable of predicting the likelihood of cancer metastasis with approximately 80% accuracy. By analyzing gene expression patterns, the AI model can assess the risk of tumor spread across multiple cancer types, including colon, breast, and lung cancers. This breakthrough offers potential for more personalized treatment strategies, enabling clinicians to identify patients who require aggressive intervention before metastasis occurs. The study challenges the traditional view of cancer as a collection of anarchic cells, proposing instead that cancer behaves as a distorted developmental process governed by biological programs reactivated through genetic and epigenetic changes. The research team focused on colon cancer cells, isolating and cloning them to observe their invasive behavior both in vitro and in mouse models. This approach allowed the identification of specific gene signatures linked to metastatic potential, which MangroveGS uses to generate its predictions. Metastasis remains the leading cause of cancer-related deaths, often detected only after the disease has spread. Despite extensive knowledge of mutations driving tumor formation, the mechanisms determining why some cells metastasize while others do not have remained elusive. The AI tool addresses this gap by integrating complex molecular data to reveal the underlying biological logic of cancer spread. Its application could transform clinical decision-making by distinguishing patients who might benefit from intensified therapies from those for whom less aggressive treatment is appropriate. Beyond clinical implications, the findings open new avenues for research into therapeutic targets aimed at halting metastasis. By decoding the molecular identity of metastatic cells, scientists may develop interventions that disrupt the biological programs enabling tumor dissemination. The study, published in Cell Reports, marks a significant advance in understanding cancer progression and highlights the growing role of AI in precision oncology.