The pharmaceuticals firm GSK has struck a five-year partnership with King’s College London to use artificial intelligence to develop personalised treatments for cancer by investigating the role played by genetics in the disease. The tie-up, which involves 10 of the drug maker’s artificial intelligence experts working with 10 oncology specialists from King’s across their labs, will use computing to “play chess with cancer”, working out why only a fifth of patients respond well to immuno-oncology treatments. Dr Kim Branson, the global head of artificial intelligence and machine learning at GSK, said only 20% of patients respond well to the new oncology drugs that harness the body’s immune system to fight cancer. “Sometimes it works like a game buster … and it wipes out the cancer. We’d like that to work all the time. This could be transformative,” Branson said. The partnership will use GSK’s cancer drugs to start with and initially focus on solid cancers such as thoracic malignancies, gastrointestinal and women’s cancers. “Hopefully we’ll create a framework that other people can contribute to,” Branson said. GSK and other large drug makers have been investing in AI to mine the vast quantities of data available to develop new medicines, pinpoint why some people are susceptible to certain diseases, and improve and personalise patient care. AI uses algorithms to carry out tasks, with computers learning through repetitive processes rather than instruction from humans. The team will use a 3D cellular model of a patient’s disease to study how tumour cells from the patients undergoing treatment interact with immune cells. “What if we could play chess with the cancer?” Branson said. “Cancer is a tricky thing. You treat with X, then you see resistance. The tumour says, ‘You do that, I’m going to respond with this.’ We’re using the predictive power of AI to think of potential strategies to outmanoeuvre disease. Our partnership with King’s can make this a possibility.” The team will monitor for dynamic biomarkers – molecules found in blood, other body fluids or tissues that are a sign of disease – that can predict resistance during treatment or a later relapse. The research partnership is based on a novel machine learning model that integrates multimodal data, genetic and molecular traits, tumour location, images and biomarker blood tests. Prof Tony Ng, head of the comprehensive cancer centre at King’s, said that in general half of cancer patients who were clinically diagnosed to have advanced but operable cancers came back within one to two years after treatment, such as chemotherapy, when it was discovered that the cancer had spread to other parts of the body. To identify those at high risk, the team will create a “digital biological twin” of the patient, to test multiple drugs, and multiple doses, at multiple time points. “We are linking up the patient with the twin and can immediately feed back info to the clinical trial or clinical management algorithms,” Ng said. “The biological twin will not only tell us this person has a high risk, but also what we as oncologists do about it.” Ng added that different parameters besides genomics can be looked at within the twin, such as whether the immune system is suppressed through contact with cancer cells (quantified by new imaging methods), to develop a “multimodal monitoring tool”. Over the five years, the team hope to create specialist equipment. Branson said the partnership could, if necessary, use the UK’s most powerful supercomputer, developed by the US-based firm Nvidia, which became operational in July. The Cambridge-1 deploys AI methods and is available to a range of organisations, including GSK and King’s.