Glioblastoma does not survive by acting alone.

The aggressive brain cancer surrounds itself with immune cells that would normally protect the body but are instead persuaded to support the tumour. They can weaken the immune response, encourage tumour growth and make treatment less effective.

Now, researchers from King’s College London and McMaster University in Canada have developed an experimental CAR-T cell therapy designed to attack both the cancer and the immune cells helping to protect it.

In preclinical models, including models created using human patient tumours, the treatment eliminated detectable cancer and produced long-term disease control. The research was published in Nature on 1 July 2026.

The therapy has not yet been tested in patients, so it should not be described as a cure. However, it presents a promising new way of approaching one of the hardest cancers to treat.

Why is glioblastoma so difficult to treat?

Glioblastoma is the most aggressive form of primary brain cancer in adults.

Rather than growing as one clearly defined mass, it can spread through the surrounding brain tissue in thin, branching patterns. Surgeons may remove the visible tumour, but microscopic cancer cells can remain hidden nearby.

Removing every affected cell is often impossible without damaging healthy parts of the brain. Surgery is therefore usually followed by radiotherapy and chemotherapy, but the cancer frequently returns.

Glioblastoma tumours are also highly varied. Different cancer cells within the same tumour may carry different biological markers. A treatment that destroys one group can leave another behind, allowing the disease to rebuild.

Average survival following diagnosis remains around 12 to 18 months, according to King’s College London. This poor outlook has created an urgent need for treatments that work differently from existing options.

How does CAR-T cell therapy work?

CAR-T therapy uses the body’s own immune cells to recognise and attack cancer.

T cells are collected and genetically modified in a laboratory. Scientists add a chimeric antigen receptor, or CAR, which allows the cells to identify a chosen marker on the surface of cancer cells.

The modified cells are multiplied before being returned to the patient, where they search for cells carrying that marker.

CAR-T therapies have already produced major results in some blood cancers. Solid tumours such as glioblastoma are much more difficult.

Their cells can vary widely, lose the marker being targeted or create an environment that prevents immune cells from functioning properly. Glioblastoma combines all of these problems with the additional challenge of treating cancer inside the brain.

Targeting the tumour and its protection

The researchers identified a protein called GPNMB that appeared in two important places.

It was found on glioblastoma cells, making those cells a potential target for CAR-T therapy. It also appeared on certain tumour-associated macrophages.

Macrophages are immune cells that normally respond to infection, remove damaged cells and help repair tissue. Inside a tumour, however, some can be manipulated into suppressing immune attacks and supporting cancer growth.

The researchers therefore developed CAR-T cells that recognised GPNMB.

Instead of attacking only the cancer cells, the therapy also targeted some of the immune cells creating a protective environment around the tumour. In the preclinical models, this two-pronged approach produced long-term tumour control.

This is what makes the research particularly interesting.

Many cancer treatments focus directly on malignant cells. However, a tumour is not simply an isolated collection of cancer cells. It exists within a wider biological system containing immune cells, blood vessels and surrounding tissue.

By disrupting both the tumour and part of the system supporting it, the new therapy may reduce the cancer’s ability to recover.

What happens next?

The findings are encouraging, but there is still a large gap between success in preclinical research and an effective treatment for patients.

Further studies will need to examine the safety of targeting GPNMB, particularly because proteins found on cancer cells can sometimes appear in healthy tissues too.

Researchers must also determine how the CAR-T cells should be delivered, what dose should be used and whether the treatment can work safely inside the human brain.

Early clinical trials would focus first on safety and tolerability. Only later could researchers establish whether the treatment improves survival or prevents the cancer from returning.

The work also shows why collaboration matters in life sciences. Professor Sheila Singh holds roles at both King’s College London and McMaster University, connecting research teams and clinical expertise across the UK and Canada.

At King’s, cancer research is linked closely with clinicians at Guy’s and St Thomas’ NHS Foundation Trust. These relationships between universities, hospitals and specialist researchers are essential when moving an idea from the laboratory towards patient care.

A promising new direction

Cancer research is often described using the word “breakthrough”, but the reality is usually more gradual.

This study has not produced a new treatment available to patients. Many therapies that work in laboratory models do not succeed during human trials.

What it has produced is a strong scientific idea.

Glioblastoma has repeatedly resisted treatments that attack one target at a time. By targeting the tumour and part of its immune protection simultaneously, researchers have revealed a possible weakness in an especially difficult disease.

Whether that weakness can be targeted safely in people is the next major question.

For the wider life sciences community, the study is another example of how a deeper understanding of cancer’s surrounding environment could shape the next generation of treatments.