Researchers have identified a previously unknown pathway helping glioblastoma grow and invade the brain, uncovering a new therapeutic target in one of the deadliest cancers.
A protein best known for helping tumours hide from the immune system has now been implicated as an active promoter of glioblastoma growth.
Australian researchers have uncovered a previously unknown pathway that appears to fuel the cancer’s aggressive spread through the brain.
The study, led by scientists at the Centre for Cancer Biology in Adelaide and published in the Proceedings of the National Academy of Sciences (PNAS), found that CD47 does far more than act as a “don’t eat me” signal protecting cancer cells from immune attack.
Instead, researchers showed the protein directly drives tumour cell proliferation, migration and invasion – independent of immune system activity.
Glioblastoma remains one of the deadliest cancers, with median survival typically less than 18 months despite surgery, radiotherapy and chemotherapy. Recurrence is almost universal.
Senior author Dr Nirmal Robinson said the findings challenge existing assumptions about the role of CD47 in cancer biology.
“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,” he said.
“What we’ve discovered is that CD47 is doing much more than that; it’s actually driving the cancer’s ability to spread and grow.”
The team found CD47 was highly concentrated at the invasive edges of glioblastoma tumours, where cancer cells infiltrate healthy brain tissue. Higher levels of the protein were also associated with poorer patient survival.
Using laboratory experiments and animal models, the researchers demonstrated that removing or blocking CD47 sharply reduced tumour cell growth and invasion. Tumours lacking the protein grew more slowly and, in some models, survival time nearly doubled.
Importantly, the effects persisted even in the absence of immune cells, confirming CD47 was exerting tumour-promoting effects directly within the cancer itself.
The researchers also identified a downstream partner protein, ROBO2, which appears central to the process.
Lead researcher Dr Ruhi Polara said CD47 effectively protects ROBO2 from cellular destruction.
“Essentially, CD47 is shielding ROBO2, allowing it to accumulate and drive tumour progression,” she said.
“When we remove CD47, ROBO2 is degraded, and the cancer cells lose their ability to grow and invade effectively.”
Related
The mechanism hinges on a third protein, ITCH, which would normally tag ROBO2 for breakdown. CD47 prevents that process by sequestering ITCH, stabilising ROBO2 and sustaining tumour progression.
Researchers described the newly identified CD47-ITCH-ROBO2 axis as a significant advance in understanding glioblastoma biology and a possible therapeutic target.
While anti-CD47 therapies are already in clinical development for several cancers, they have so far delivered limited results in glioblastoma. The authors suggest directly targeting the CD47-ROBO2 pathway, or interfering with ROBO2 stabilisation, may prove more effective.
“By understanding this mechanism, we now have new targets to explore,” Dr Polara said.
“This could lead to the development of therapies that specifically block the tumour’s ability to spread, which is one of the biggest challenges in treating glioblastoma.”
The researchers said their study also highlighted the importance of looking beyond the immune system when developing cancer treatments.
“This work changes how we think about CD47,” Dr Robinson said.
“It’s not just an immune checkpoint; it’s a central regulator of tumour biology in its own right.”
The researchers said further studies were needed to translate these findings into clinical treatments, but the discovery represented a significant step forward in the fight against one of the most devastating cancers.
Proceedings of the National Academy of Sciences (PNASS), March 2026
