New AI tool predicts cancer spread with surprising accuracy

A new AI tool can predict which cancers will spread—before it’s too late.

· ScienceDaily
Source:Université de Genève
Summary:Researchers have discovered that cancer spread isn’t random—it follows a kind of biological “program.” By studying colon tumor cells, they identified gene patterns that signal whether a cancer is likely to metastasize. Their AI model, MangroveGS, can predict this risk with about 80% accuracy and even works across multiple cancer types. This could transform how doctors decide who needs aggressive treatment and who doesn’t.
Group of human colon cancer cells with invasive behaviour. Cell nuclei are in yellow and cell bodies in red. The finger-like protrusions of invasive cells are on the upper right region. Credit: © Ariel Ruiz i Altaba, UNIGE

Why do some tumors spread while others remain localized? Scientists still do not fully understand what controls a cancer cell's ability to metastasize, but answering this question is essential for improving patient care. Researchers at the University of Geneva (UNIGE) studied cells from colon cancer and identified key factors that influence whether a tumor is likely to spread. They also uncovered specific gene expression patterns that can be used to estimate that risk.

Building on these findings, the team developed an artificial intelligence tool (MangroveGS) that converts these genetic signals into highly reliable predictions across multiple cancer types. The study, published in Cell Reports, could lead to more personalized treatments and help uncover new therapeutic targets.

Cancer as a Distorted Development Process

"The origin of cancer is often attributed to 'anarchic cells'," explains Ariel Ruiz i Altaba, professor in the Department of Genetic Medicine and Development at the UNIGE Faculty of Medicine, who led the study. "However, cancer should rather be understood as a distorted form of development." Genetic and epigenetic changes can reactivate biological programs that are normally turned off after early development, ultimately driving tumor formation.

Rather than being random, cancer appears to follow structured biological rules. "The challenge is therefore to find the keys to understanding its logic and form. And, in the case of metastases, to identify the characteristics of the cells that will separate from the tumor to create another one elsewhere in the body."

Tracking Metastatic Cancer Cells

Metastasis is responsible for most cancer deaths, especially in colon, breast, and lung cancers. By the time cancer cells are detected circulating in the blood or lymphatic system, the disease has often already begun to spread. Although scientists understand many of the mutations that lead to tumor formation, no single genetic change explains why some cells break away and migrate while others remain in place.

"The difficulty lies in being able to determine the complete molecular identity of a cell - an analysis that destroys it - while observing its function, which requires it to remain alive," explains Professor Ruiz i Altaba. To overcome this, the researchers isolated, cloned, and grew tumor cells in the lab. "These clones were then evaluated in vitro and in a mouse model to observe their ability to migrate through a real biological filter and generate metastases," adds Arwen Conod.

Gene Signatures Linked to Cancer Spread

The team analyzed the activity of hundreds of genes in about thirty cell clones taken from two primary colon tumors. This revealed clear gene expression patterns that closely matched each cell's ability to move and spread. Importantly, metastatic potential was not determined by a single cell's profile, but by how groups of related cancer cells interact with each other.

AI Tool Predicts Metastasis Risk

The researchers integrated these gene signatures into an artificial intelligence system. "The great novelty of our tool, called 'Mangrove Gene Signatures (MangroveGS)', is that it exploits dozens, even hundreds, of gene signatures. This makes it particularly resistant to individual variations," explains Aravind Srinivasan.

After training, the model was able to predict metastasis and colon cancer recurrence with nearly 80% accuracy, outperforming existing methods. The same gene signatures derived from colon cancer also proved useful in predicting metastatic risk in other cancers, including stomach, lung, and breast cancer.

Toward More Personalized Cancer Care

MangroveGS can work directly with tumor samples collected in hospitals. Cells are analyzed, their RNA is sequenced, and a metastasis risk score is quickly generated and shared securely with doctors and patients through an encrypted platform.

"This information will prevent the overtreatment of low-risk patients, thereby limiting side effects and unnecessary costs, while intensifying the monitoring and treatment of those at high risk," says Ariel Ruiz i Altaba. "It also offers the possibility of optimizing the selection of participants in clinical trials, reducing the number of volunteers required, increasing the statistical power of studies, and providing therapeutic benefits to the patients who need it most."

Related Terms

CancerCancer is a group of diseases in which cells are aggressive (grow and divide without respect to normal limits), invasive (invade and destroy adjacent tissues), and/or metastatic (spread to other locations in the body). These three malignant properties of cancers differentiate them from benign tumors, which are self-limited in their growth and do not invade or metastasize (although some benign tumor types are capable of becoming malignant). Cancer may affect people at all ages, even fetuses, but risk for the more common varieties tends to increase with age. Cancer causes about 13% of all deaths. Apart from people, forms of cancer may affect animals and plants.GeneA gene is a locatable region of genomic sequence, corresponding to a unit of inheritance, which is associated with regulatory regions, transcribed regions and/or other functional sequence regions. The physical development and phenotype of organisms can be thought of as a product of genes interacting with each other and with the environment, and genes can be considered as units of inheritance. A concise definition of gene taking into account complex patterns of regulation and transcription, genic conservation and non-coding RNA genes, has been proposed by Gerstein et al. "A gene is a union of genomic sequences encoding a coherent set of potentially overlapping functional products."Gene therapyGene therapy is the insertion of genes into an individual's cells and tissues to treat a disease, and hereditary diseases in which a defective mutant allele is replaced with a functional one. Although the technology is still in its infancy, it has been used with some success. Antisense therapy is not strictly a form of gene therapy, but is a genetically-mediated therapy and is often considered together with other methods. In most gene therapy studies, a "normal" gene is inserted into the genome to replace an "abnormal," disease-causing gene. A carrier called a vector must be used to deliver the therapeutic gene to the patient's target cells. Currently, the most common type of vectors are viruses that have been genetically altered to carry normal human DNA. Viruses have evolved a way of encapsulating and delivering their genes to human cells in a pathogenic manner. Scientists have tried to harness this ability by manipulating the viral genome to remove disease-causing genes and insert therapeutic ones.Monoclonal antibody therapyMonoclonal antibody therapy is the use of monoclonal antibodies (or mAb) to specifically target cells. The main objective is stimulating the patient's immune system to attack the malignant tumor cells and the prevention of tumor growth by blocking specific cell receptors. Variations exist within this treatment, e.g.radioimmunotherapy a radioactive dose directly to the target cell, and lethal chemical doses to the target.AlleleAn allele is a viable DNA (deoxyribonucleic acid) coding that occupies a given locus (position) on a chromosome. Usually alleles are sequences that code for a gene, but sometimes the term is used to refer to a non-gene sequence. An individual's genotype for that gene is the set of alleles it happens to possess. In a diploid organism, one that has two copies of each chromosome, two alleles make up the individual's genotype.Adult stem cellAdult stem cells are undifferentiated cells found throughout the body that divide to replenish dying cells and regenerate damaged tissues. Also known as somatic  stem cells, they can be found in children, as well as adults.Antibiotic resistanceAntibiotic resistance is the ability of a microorganism to withstand the effects of an antibiotic. It is a specific type of drug resistance. Antibiotic resistance evolves naturally via natural selection through random mutation, but it could also be engineered by applying an evolutionary stress on a population. Once such a gene is generated, bacteria can then transfer the genetic information in a horizontal fashion (between individuals) by plasmid exchange. If a bacterium carries several resistance genes, it is called multiresistant or, informally, a superbug.Embryonic stem cellEmbryonic stem cells (ESCs) are stem cells derived from the undifferentiated inner mass cells of a human embryo. Embryonic stem cells are pluripotent, meaning they are able to grow (i.e. differentiate) into all derivatives of the three primary germ layers: ectoderm, endoderm and mesoderm. In other words, they can develop into each of the more than 200 cell types of the adult body as long as they are specified to do so.