Brain cancer is the kind of headline that makes your stomach drop. Most people do not spend their days reading research papers, but when a study touches something as frightening as glioblastoma, it lands differently. It is the most aggressive form of brain cancer and it is notoriously hard to treat. Even with surgery, radiation and chemotherapy, it almost always comes back.
A new Australian study has not found a cure, but it has uncovered something researchers have been missing: a protein that does more than hide tumours from the immune system. It may also help glioblastoma spread through the brain, which is one of the biggest reasons this cancer is so difficult to control.
The discovery, led by Adelaide University researchers at the Centre for Cancer Biology and published in PNAS, points to a new pathway that could eventually guide smarter treatments.
What is glioblastoma?
Glioblastoma is one of the most aggressive brain cancers and most patients survive under 18 months after diagnosis. Treatments can slow it down, but it often returns. The new research, led by Adelaide University scientists, points to a “hidden driver” inside tumour cells that may help explain why.
When glioblastoma returns, it is often because tiny clusters of tumour cells have already moved into nearby healthy brain tissue. Those cells can sit beyond the reach of surgery and later grow into new tumours.
That is why researchers pay close attention to the tumour’s “edges”, the invading front where the cancer pushes outward. In this study, the team found a protein called CD47 is especially concentrated in those invasive zones and higher CD47 levels linked with poorer survival outcomes.
In other words, the most dangerous part of the tumour seems to have more of the very thing that helps it keep moving.
CD47 was already known, just not like this
CD47 has been studied for years because of its role in immune evasion. Scientists often describe it as a “don’t eat me” signal that helps cancer cells avoid being destroyed by the immune system.
“We’ve known for some time that CD47 acts as a kind of ‘don’t eat me’ signal that helps cancer cells hide from the immune system,” says senior author of the study Dr Nirmal Robinson.
The twist is what the team found next.
“What surprised us was finding that CD47 is also acting inside glioblastoma cells, directly driving their ability to grow, move and invade, even when immune cells aren’t present,” says lead researcher Dr Ruhi Polara.
That detail matters because it suggests CD47 is not only a shield. It is also an engine.
Why that changes the story
Plenty of cancer treatments target the immune system. Some experimental therapies already aim at blocking CD47 so the immune system can recognise and attack cancer.
But glioblastoma is a unique problem. The brain has its own immune environment and tumours can behave differently there. If CD47 is helping tumours grow and spread even when immune cells are not involved, then immune-based strategies alone may not be enough.
“Our findings suggest researchers may need to rethink CD47 as a target, not just as an immune checkpoint, but as a protein that is actively driving tumour growth and invasion from within the cancer cell itself,” Dr Polara says.
The new pathway
The study identified a chain reaction inside tumour cells involving CD47 and two other proteins: ITCH and ROBO2.
Here is the simplest version:
- ROBO2 helps glioblastoma cells behave more aggressively, including spreading through brain tissue
- ITCH is a “clean-up” protein that normally tags ROBO2 for breakdown
- CD47 gets in the way, stopping ITCH from doing that job
- That means ROBO2 builds up, which can help the tumour keep growing and invading
“The CD47–ITCH–ROBO2 pathway explains how CD47 facilitates tumour progression,” Dr Polara says. “In plain terms, CD47 keeps ROBO2 stable by limiting ITCH’s ability to break it down.”
This is one of the reasons the finding has traction. It gives researchers a clearer map of what is happening inside the cancer cell, not just around it.
Why this cancer is so hard to treat
Glioblastoma is not only aggressive. It is also sneaky.
The cells at the invasive edge are the ones most likely to slip beyond what surgeons can remove. That is why relapse is so common, even after major treatment.
“The invasive edge of glioblastoma tumours comprises of highly aggressive glioblastoma cells, which not only prevent complete removal of the tumour at the time of surgery, but also enables the tumour to relapse after several months of surgery,” Dr Polara says.
And importantly, CD47 appears to be active in that exact zone, even when immune cells are scarce. That suggests the protein may help tumours spread through the brain in a way that does not rely on immune evasion at all.
What this could mean for future treatments
The research also helps explain why CD47-targeting therapies have had limited success in glioblastoma so far. Many of those approaches focus on the immune system angle.
The new finding suggests the next phase may need to target tumour behaviour directly, for example by disrupting the CD47–ROBO2 relationship, or going after ROBO2 itself.
That could also matter for side effects. CD47 exists on healthy cells too, which is part of why blocking it can cause toxicity.
“Blocking CD47 to suppress its immune evasion functions has been shown to induce toxicity in patients as CD47 blockade not only targets tumour cells but also affects healthy cells that express CD47,” Dr Polara says.
The study proposes that targeting ROBO2 alone, or combining a ROBO2 approach with lower-dose CD47 strategies, could be a safer and more effective path. That still needs testing, but it is a more precise idea than a blanket “block CD47 and hope” approach.
What patients and families should take from this now
This is not a new treatment today. It is not a cure. But it is a meaningful step in understanding why glioblastoma is so relentless.
“For patients and families living with glioblastoma right now, the most important takeaway is not that this discovery leads to an immediate new treatment, but that it deepens our understanding of why this cancer is so aggressive and difficult to control,” Dr Polara says.
And it offers something research families often want: a clearer direction for what might come next.
“The key advance here is that we have identified a clearer biological target, which helps focus drug development efforts and increases the chances that future treatments will be more effective than those currently available,” she says.
The bottom line
This research reframes CD47 as more than an immune escape trick. It may also be helping glioblastoma do what it does best: spread.
“This work changes how we think about CD47,” Dr Robinson says. “It’s not just an immune checkpoint, it’s a central regulator of tumour biology in its own right.”
For now, the practical impact is knowledge. A stronger explanation for why glioblastoma returns and a new set of targets researchers can test. For a disease where progress feels slow and stakes feel impossibly high, that still matters.
Newsletter
