Artificial Intelligence (AI) can predict whether adult brain cancer patients will survive more than eight months after receiving radiotherapy.
Using AI to successfully predict patient outcomes will allow clinicians to be better informed in planning the next stage of treatment and refer patients to potentially life-saving treatment more quickly.
This is the first use of artificial intelligence to predict short- and long-term survivors within eight months of radiotherapy.
The paper recently published in Neuro-Oncology shows how researchers from King’s College London built a deep learning model to enable them to more reliably and accurately predict outcomes for patients with adult primary brain cancer.
Glioblastoma is a difficult cancer to treat, with only one in four patients surviving more than a year after diagnosis. The researchers applied deep learning – a type of artificial intelligence – to predict whether glioblastoma patients would survive eight months after receiving radiation therapy. Eight months is usually the time it takes to complete a typical course of routine chemotherapy that usually follows radiation therapy.
Currently, patients are regularly and regularly checked to see if the chemotherapy is working. But this means that some patients have ineffective chemotherapy that would not save their lives and will suffer harmful side effects.
Instead, by giving an instant and accurate prediction from a routine MRI, AI allows doctors to identify patients who would not benefit from chemotherapy to try a different treatment or start an experimental treatment in a clinical trial.
This study was motivated by a clinically focused and critical research question about aggressive brain tumors and delivered by leveraging cutting-edge artificial intelligence. Although less common than other cancers, the devastation is disproportionate with a two-year survival rate of 18%.
Dr Thomas Booth, Reader in Neuroimaging at King’s College London and Consultant in Neurology at King’s College Hospital NHS Foundation Trust
Alysha Chelliah, PhD researcher from King’s College London, said: “We applied deep learning to predict whether glioblastoma patients would survive in the first eight months after completing radiotherapy. The AI model showed improved performance when first trained to detect abnormalities in 10,000 brain MRIs. This approach aims to improve the ability to identify patients who need early second-line treatment or enrollment in clinical trials, compared to those who show an initial response to treatment.”
The researchers trained the AI on a dataset of 10,000 scans from all types of brain cancer patients.
Dr Thomas Booth said: “Feedback from all patients and clinicians at the start of the study meant we wanted to address the unmet need to improve outcomes for the large proportion of patients on modified treatment – usually shorter duration and lower dose radiotherapy if chemotherapy is not effective – as are the minority of patients who can tolerate the ‘optimal’ treatment. Almost all previous research has looked only at the latter group of patients.
“We’ve also narrowed down a thorny issue: after radiation therapy, follow-up brain scan findings are often nonspecific, and oncologists can’t be sure whether a treatment is working or failing.
“Instead of trying to interpret every non-specific brain scan, we simply looked at a routine scan after radiation therapy and gave an accurate prediction using artificial intelligence to answer a simple question: which patients will not survive the next 8 months? AI has been able to give us an immediate and accurate prediction which means clinicians can empower patients to make choices about their treatment.”
Dr Booth added: “We would be delighted if the cancer research community is now using our AI tool to see improved outcomes for patients who would not benefit from the usual course of chemotherapy.”
Commenting on how the work of Dr Booth and his team at King’s College London has supported the first human brain tumor research, as supported by the charity brainstrust, Dr Helen Bulbeck, Director of Services and Policy at brainstrust he said:
“This is exciting and fundamental research for people living with glioblastoma for two reasons. At its simplest level it shows how artificial intelligence can be used to benefit patients. More importantly, however, it empowers patients and their caregivers to make choices about the clinical pathway and gives back control at a time when so much control has been lost.Patients will be able to make informed decisions about treatment options and be able to plan how they want to spend the time they have left, so they can live their best day, every day.”
Dr Michele Afif, Chief Executive at The Brain Tumor Charity added: “Using artificial intelligence to assess and predict response to radiotherapy much earlier in a patient’s treatment for glioblastoma is a hugely important step towards tackling this of the notoriously difficult-to-treat disease. .
At The Brain Tumor Charity, we welcome this important advance which could lead to more informed discussions at an early enough point in a patient’s treatment to consider important alternatives such as clinical trials.
We look forward to seeing how this exciting research develops as it is validated for wider use as a tool to improve care for those diagnosed with a brain tumor.”
The study involved a collaboration of 11 neuro-oncology centers from across the UK.
Source:
Journal Reference:
Chelliah, A., et al. (2024). Glioblastoma and Radiotherapy: a multicenter study of artificial intelligence for survival prediction from MRI (GRASP study). Neuro-Oncology. doi.org/10.1093/neuonc/noae017.