Researchers have developed a new preclinical platform for studying aggressive childhood brain tumours, using advanced imaging to show how closely experimental models mirror the disease seen in patients.
Researchers at The Institute of Cancer Research, London, led work to develop a large panel of patient-derived mouse models of paediatric-type diffuse high-grade glioma (PDHGG) – a group of brain tumours with poor survival rates. Using magnetic resonance imaging (MRI), the research team demonstrated that these models reproduce key biological and radiological features of the human disease.
The study, published in Neuro-Oncology Advances and primarily funded by Cancer Research UK, provides a new framework for testing potential therapies and monitoring tumour response to closely reflect what happens in the clinic – helping researchers identify the most promising treatments for patients more quickly. Additional funding was received by CRIS Cancer Foundation and from a number of our family charity partners, including Abbie's Army, The Rudy A Menon Foundation, The Ollie Young Foundation, Lucas' Legacy and the Doing It For Daniel Foundation.
Building better models
PDHGGs, aggressive and fast-growing brain cancers, are among the leading causes of cancer-related morbidity and mortality in children and young adults. In most cases, median survival is only 9–18 months, with two- and five-year survival rates as low as 10 and two per cent, respectively.
Despite decades of research, outcomes have largely remained unchanged. This is primarily because these tumours are biologically distinct from other adult brain cancers with similar presentation, and they are difficult to model accurately in the laboratory.
To address this, the research team established 35 patient-derived, site-specific tumour models by implanting tumour material into the equivalent locations in the brains of mice. The models represent a range of PDHGG subtypes, including diffuse midline glioma (DMG) and diffuse hemispheric glioma (DHG).
Survival times varied widely across the models, reflecting the diversity seen in patients. While some tumours grew slowly over many months, others progressed more rapidly – an important feature to observe for testing new treatments.
Using MRI to understand tumour behaviour
Rather than relying solely on tissue analysis at the end of the experiment the researchers used non-invasive MRI to track tumour growth and behaviour over time. This approach allowed them to assess the size and structure of the tumour and the integrity of the blood-brain barrier – all vital factors in treatment response.
In the MRI images, the tumours varied in appearance, ranging from widespread growth to well-defined masses.
A key finding was that most tumours did not light up on MRI scans after a contrast agent was given, showing that the blood-brain barrier was still intact and blocking substances from entering the tumour – a major obstacle for drug delivery in patients. This was specifically true for tumours in the brainstem – the area connecting the brain to the spinal cord – mirroring clinical observations of being harder to treat and emphasising the importance of testing therapies in models that reflect this challenge.
The team also identified differences in water diffusion within tumours, using a technique called diffusion-weighted imaging, which generates images by mapping the random movement of water molecules in tissues. Brainstem tumours had higher apparent diffusion values than hemispheric tumours, a pattern also seen in patient scans, further confirming that the models closely reflect how these cancers behave in people.
Further examinations
The research team also examined whether tumours changed as they were re-implanted into new mice, a common step in maintaining experimental models. While some tumours grew faster when re-implanted, most retained the characteristics seen with MRI, suggesting that their fundamental biological features and appearance remained stable.
In a further set of experiments, tumours grown from cells cultured in three-dimensional conditions showed shorter survival and higher proliferation than those grown in traditional two-dimensional cultures. This demonstrates how laboratory growth conditions can influence behaviour and reinforces the value of carefully designed preclinical systems.
Informative preclinical trials
First author Dr Jessica Boult, Staff Scientist in the Pre-Clinical MRI Group at The Institute of Cancer Research (ICR), said: “This research lays important groundwork for future therapy development. If we want to improve outcomes for children and young adults with these tumours, we need models that genuinely reflect the disease – not just at the molecular level but in how tumours grow, spread and respond to treatment. Embedding advanced imaging into these models allows us to do that.”
By combining patient-derived tumours with clinically relevant MRI techniques, the platform enables researchers to evaluate whether new drugs reach the tumour, alter its biology and produce meaningful changes over time.
The study provides a valuable shared resource for the cancer research community, offering a realistic testing ground for much-needed treatments for PDHGGs. It also reinforces the role of imaging as a bridge between laboratory research and preclinical trials, ensuring that promising treatments are assessed in ways that best predict patient benefit.
Senior author Professor Simon Robinson, Group Leader of the Pre-Clinical MRI Group, said: “For children and families affected by these aggressive brain tumours, the ultimate aim is to accelerate the development of treatments that can finally improve survival. This imaging-led approach brings us a step closer to that goal.”
The ICR has been instrumental in driving progress in glioma. Find out more about how we have led the way in childhood cancer research over the decades.