[Science] Mini organs grown from tumour cells can help us choose the best chemo – AI

By Jessica Hamzelou We can now test how cancer cells respond to a drug before people begin treatmentMAURIZIO DE ANGELIS/SCIENCE PHOTO LIBRARY Miniature organoids grown from a person’s cancer cells could help to predict whether or not they will respond to some types of chemotherapy. With improvement, the organoids could help further personalise cancer treatment, say the researchers behind the work. At the moment, it is difficult to know whether a chemotherapy will work for an individual – while the drugs are often lifesaving, in some cases they can trigger horrendous side effects without having any benefit. To better predict if a treatment will work or not, teams around the world have been working to develop personalised organoids – using clumps of cells biopsied from a person’s tumour. The idea is that the organoid can serve as a laboratory model for the person with cancer, and that drugs that kill cells in the organoid are more likely to effectively treat their tumour. Advertisement Emile Voest at the Netherlands Cancer Institute and his colleagues tested this theory by attempting to grow organoids from 67 biopsies taken from 61 people with colorectal cancer. The organoids take three weeks to grow. Due to difficulties in obtaining cancer cells, contamination issues and problems with evaluating individuals’ response to treatment, the team were only able to test 35 organoids. Correct prediction The team first gave a subset of the organoids a drug called irinotecan for six days. They found that the drug seemed to work similarly in the organoids and in the individuals the cells were taken from. Tests on the organoids correctly predicted how eight out of 10 people responded to treatment. But the organoids failed to predict how well a combination treatment of irinotecan and a drug called fluorouracil worked. This might be because they are too simplistic a model. A person’s immune system plays an important role in how they respond to drugs, and the organoids don’t have one, say the study authors. Using a different method to study the cells might solve some of the problems, says Helen Abud at Monash University, Australia. “Overall, this study does provide considerable hope that organoids could be used to predict patient drug response,” she says. The team is confident that, with some improvements, organoids could soon be used to help tell which chemotherapy might be best for an individual. Elizabeth Vincan at the University of Melbourne, Australia, agrees. “[The researchers] demonstrate that patient-derived colon cancer organoids do indeed predict patient treatment outcome for irinotecan-based chemotherapy,” she says. “Thus, we can truly personalise therapy and avoid unnecessary treatment.” Journal reference: Science Translational Medicine, DOI: 10.1126/scitranslmed.aay2574 More on these topics: cancer health personalised medicine