An artificial intelligence (AI) project has demonstrated progress towards developing personalised prostate cancer treatments, researchers reported.
A team from University College London (UCL) used an AI tool called ArteraAI to analyse data from a previous trial measuring the effectiveness of different prostate cancer treatment regimens. With the help of ArteraAI, the team gained new insight into which treatments were most effective for individual men with the disease.
Gerhardt Attard, a professor from UCL who led the project, said: "This ability to better predict the behaviour of cancer from routinely performed tests could be a game-changer."
Developing Precision Treatment
Researchers on the STRATOSPHere project are analysing the cellular mechanics of different types of prostate cancer from thousands of blood and tumour samples donated by men during the previous trial, called STAMPEDE.
The aim is to try to speed up the development of precision treatment so that men with advanced prostate cancer could eventually be offered medicine matched to the specific biology of their cancer.
The STAMPEDE trial tested the effectiveness of different treatments combined with hormone therapy for men with aggressive prostate cancer. The goal was to help patients live longer and feel healthier by establishing which combination was the most successful. Although the trial successfully identified treatment regimens that worked best on average, it lacked the ability to predict which specific treatments were most effective for individual patients.
However, the latest results, presented at the European Society for Medical Oncology Congress in Madrid, suggested a significant leap in identifying men who were responding to current treatment. Also, men who weren't responding could be identified for possible participation in clinical trials to test more intensive treatments designed specifically for them.
Based on their findings, the researchers concluded that ArteraAI's model was better at predicting outcomes for men with advanced prostate cancer than existing prognostic methods. The tool was also able to use biopsy slide images of men with localised disease to identify features suggesting a higher risk of their cancer progressing to an advanced stage.
Closer to Personalised Treatments
Professor Attard explained: "With the help of models derived from artificial intelligence, we have been able to obtain new information from materials that we already have in the course of standard clinical practice, such as tissue samples. This new knowledge can enable us to accurately predict how specific cancers, and therefore which men, are likely to respond to treatment combinations physicians now routinely use, thanks to the crucial work of STAMPEDE."
Prostate Cancer UK (PCUK) and Movember, a men's health charity, donated £1.3 million towards supporting this research.
Dr Matthew Hobbs, PCUK's director of research, commented: "I'm particularly excited to see the opportunity brought by combining excellent UK clinical research, careful and deliberate research funding, and innovative AI technology. This combination now sets the path for the next generation of clinical trials for men with very aggressive prostate cancer, bringing us closer to effective, personalised treatments being the norm for every man, sooner than we thought possible."
Dr Sarah Hsiao, director of biomedical research and impact at Movember, said the charity was proud to have supported the innovative research, which "employs state-of-the-art technology to develop a tool that can assist men with prostate cancer and their doctors to make the most informed treatment choices." She described the latest findings as "a significant step towards the concept of providing the right treatment for the right men at the right time."
Professor Nick James, a consultant clinical oncologist and chief investigator on the STAMPEDE trial, told Medscape News UK: "This work shows the value of pairing the incredibly rich data and samples from STAMPEDE with artificial intelligence to transform our ability to predict which treatments will be best for each patient, depending on the unique biology of their cancer. Artificial intelligence is giving us a new edge in the fight against cancer, and I hope this work will pave the way towards personalised treatment plans which not only extend men's lives, but also grant them many extra years of living well, with fewer side effects."