Radiotherapy is a vital part of so many brain tumour and cancer treatments, but in many cases, this treatment pathway starts long before a patient makes it to a radiotherapy centre and often involves other forms of treatment working in concert before they are given the all clear.

Despite being a medical practice that is over a century old at this point, radiotherapy is still constantly developing both in terms of the types of radiation used and the increasingly precise and adaptable techniques used to ensure precision and efficacy.

As well as this, there are developments in the field of diagnostics, something vital for effective radiotherapy treatment as the earlier a tumour is spotted, the easier it can be to treat and the less likely it has grown or metastasised.

One of the most interesting reports in this regard is a King’s College London study that claims that AI can be effectively used to predict the prognosis of patients after a course of radiotherapy treatment.

Is this true? If it is, how does this affect the practice of radiotherapy in specific and cancer treatment in general?

What Does The Study Say?

The paper in question was published in Neuro-Oncology and focused on the ability to predict the short-term and long-term outcomes of cancer treatment in patients with brain cancer.

This study specifically looks at adult glioblastoma, a fast-growing and aggressive form of brain cancer that comprises 15 per cent of all brain tumours and has an average survival time after diagnosis of just over a year, with five per cent surviving five years or more.

The treatment pathway for glioblastoma is a targeted, aggressive course of radiotherapy followed by a course of chemotherapy that typically lasts eight months.

The study looked at a single MRI scan after being trained on tens of thousands of scans, and through a deep learning process examined whether a patient would reach the eight-month mark that signals the completion of chemotherapy.

The focus of the study was not necessarily to determine the exact prognosis of patients, which can be affected by a wide range of factors, but instead to see if a patient would have the chance to complete a course of chemotherapy.

The study itself had a similar level of accuracy to the current methodology or regular brain scans during the chemotherapy process itself but was achieved before the chemotherapy course had even started.

Why Does This Matter For Radiotherapy?

At present, it can be difficult to ascertain the effectiveness of cancer treatment after the radiotherapy phase, due to the nonspecific nature of brain scans taken during the course of chemotherapy.

This means that, in some cases, patients are on an eight-month course of treatment that may not necessarily help extend their life, but due to the side effects of the treatment may affect their quality of life during that time.

The combination of mental and physical distress could potentially be avoided if a single scan after radiotherapy reveals that the primary course of treatment may not necessarily be the most effective rather than waiting for a response or lack thereof to treatment.

This is beneficial for a huge number of reasons, the biggest of which is giving patients as much time as possible and agency over their own treatment decisions whilst minimising unnecessary pain and side effects.

This could include trying a different treatment or an additional round of radiotherapy or taking part in a clinical trial to see if an experimental treatment could help.

This also allows for a wider range of meaningful choices to be available to patients and caregivers to ensure that they have access to the widest range of treatments possible.

These choices could include radiotherapy, novel medications, combination treatments including lower dose radiation and alternative short-course chemotherapy and a variety of new technologies currently the subject of clinical trials.

It could also include palliative radiotherapy that does not intend to treat the condition but ensures that patients remain comfortable and can make the most of every day.

This also matters for radiotherapy as a highlight of the effects artificial intelligence has on the field, from initial diagnosis to follow-up treatments, as well as a way to distinguish signal from noise given the particularly fast-paced research into AI’s place in the wider world.

AI has been used to help diagnose cancer for years, including a modified bread checkout scanner that can detect an entire slide of microscope cells from certain microscopic markets.

As well as this, AI has been used in drug discovery, helping to speed up a process that can take decades in order to find effective treatments for patients.