NIHR supports study to predict the side effects of radiotherapy: Hamlet.rt
- 8 October 2024
- 4 min read
Oncology Translational Research Collaboration (O-TRC) supports study to predict the side effects of radiotherapy and generate better outcomes for patients.
Radiotherapy treatment is used to treat cancer but causes long-term side effects
Radiotherapy is used to treat cancer, often alongside surgery and drug therapy. Approximately half of all cancer patients receive radiotherapy during the course of their illness. During radiotherapy treatment, the highest dose of radiotherapy possible is used. This unfortunately results in long-term side effects which can continue after treatment has ended. This can include skin changes, permanent hair loss in the treated area and fatigue.
Researchers have tried to understand why radiotherapy leads to certain side effects. This is being done using prediction models. The models look at where radiation has been admitted to the body to try and understand or predict what the side effects will be. However, the usefulness of these models are limited. A person’s medical history can also impact the likelihood of certain side effects occurring. High blood pressure, diabetes, smoking and certain medications can impact response to treatment. Additionally, previous prediction models have been based on a small number of patients. This has made findings less applicable to the wider population.
Current studies to predict side effects are based on small numbers of patients
Dr Raj Jena, Principal Research Associate at the University of Cambridge, Honorary Consultant at Addenbrooke's Hospital and Chair of the NIHR Oncology TRC, highlights that to build better prediction models, they would have to be larger than the typical 150-200 patient based models published to date.
Artificial intelligence (AI) and machine learning has been recently adopted to help analyse data collected from research studies. Dr Jena sought to develop a study which will collect data directly from patients and use AI technology to analyse the findings. The study is known as Hamlet.rt (Heuristics And Machine Learning Evaluation and Testing in RadioTherapy).
The type of data collected from Hamlet.rt includes imaging and clinical assessment data. The data is collected before, during and after radiotherapy. A questionnaire will be used to ask patients about their symptoms and side effects. This data will then be used to develop an AI computer tool that can predict the likelihood and severity of a side effect occurring in a patient. This will then enable doctors to better prepare their patients who are due to start radiotherapy treatment for any long-term side effects they might encounter.
“The motivation for the Hamlet.rt study was to collect all of the radiotherapy data and the side effects data directly from patients and use AI techniques to help to answer these questions”. - Dr Jena
Hamlet.rt was originally developed in partnership with Microsoft Research and was based at one hospital in Cambridge. The aim was to recruit 500 head and neck and prostate cancer patients over a 3-year period. However, using Hamlet.rt in Cambridge only meant side effects could only be predicted for two types of cancer. It was also limited to a particular demographic and socioeconomic population. The Hamlet.rt model needed to reach more patients in as many locations around the country as possible, and include more types of cancer.
Oncology TRC supports UK-wide delivery of Hamlet.rt
Dr Jena praises the Oncology TRC for enabling Hamlet.rt to achieve these goals. The radiotherapy workstream within the Oncology TRC provided a forum for Dr Jena to highlight the study to a number of other centres.
“We recognised that we could achieve better success if we could deliver Hamlet.rt at a greater scale than we could as individual centres and the O-TRC could provide the opportunity for us to do this.” - Dr Jena
Key centres such as Manchester BRC, Leeds BRC and Oxford BRC saw the benefits of opening up the study at their own centres. With the support of the Oncology TRC, the group was successfully awarded £40,896.50 from a Cancer Research UK (CRUK) RadNet Cambridge research award, to enable collaboration with RadNet. RadNet is a UK-wide radiation research network of centres of excellence funded by CRUK that aim to tackle major challenges in radiotherapy. Hamlet.rt has now been able to open at twelve sites across the UK. The centres focus on the top four cancers treated with radiotherapy (head and neck, prostate, lung and brain).
“Without the TRC structure to pull people together, it would not have been possible to open the study at so many more sites and ultimately establish Hamlet.rt as the leading radiotherapy AI study in the UK”
Hamlet.rt is currently on track to recruit 1,500 patients by October 2024. This is compared to the original target of 500 patients when Hamlet.rt was only open at one site.
The aim is for Hamlet.rt to better predict an individual's risk of getting long-term side effects. With this, cancer patients and doctors can be better prepared before treatment starts. This will lead to safer and kinder radiotherapy.
