What Did a 2025 Glioblastoma CAR-T Trial Actually Show?

Glioblastoma is one of the most feared diagnoses in medicine: an aggressive brain cancer that almost always returns after first treatment, with very few good options once it does. So when a headline says "CAR-T therapy for glioblastoma", it understandably sparks hope — and deserves a careful, honest read. This 2025 phase 1 trial is a genuine step forward, but it is an early one, and what it does and does not show both matter ¹.

CAR-T therapy has transformed parts of blood-cancer care, but turning it against solid tumours like brain cancer has been much harder. This trial is part of a serious effort to change that — and reading it well means separating real progress from premature hope.

What question did the researchers ask?

Researchers at the University of Pennsylvania asked a focused early-stage question: can we safely deliver a new kind of CAR-T cell directly into the brain of people whose glioblastoma has come back, and is there any early sign of benefit ¹?

Two design choices make this trial interesting. First, the CAR-T cells were bivalent — engineered to recognise two targets on the cancer at once (a part of EGFR and a protein called IL-13Rα2), rather than just one. That matters because solid tumours often escape single-target therapies by simply losing the one target. Second, the cells were delivered intracerebroventricularly — infused into the fluid-filled spaces of the brain, putting the therapy close to the tumour rather than relying on cells injected into a vein to find their way in.

If the basic idea of CAR-T is new to you, our explainer on what CAR-T therapy is covers the concept first.

What did the trial find?

This was a phase 1 trial, and it's important to understand what that means: phase 1 studies are small and designed primarily to test safety and find a usable dose — not to prove a treatment works. With that framing, the findings were cautiously encouraging.

The approach was judged feasible and tolerable, and the researchers identified a maximum tolerated dose. On early signs of activity, the trial reported one patient with a partial response (meaningful tumour shrinkage) and another with durable stable disease (the cancer held steady for a meaningful time) ¹. In recurrent glioblastoma — where progression is usually rapid — even those signals are noteworthy.

But the safety findings are just as important. The therapy caused early-onset neurological toxicity, a known risk when powerful immune cells are activated next to the brain. In this trial that toxicity was manageable in a specialist setting, but it underlines that this is intensive, closely monitored hospital care.

How strong is this evidence?

This is genuinely meaningful science, but its place on the evidence ladder is low by design. A phase 1 trial in a handful of patients can tell us that an approach is safe enough to study further and worth pursuing — it cannot tell us how well it works, how long benefits last, or how it compares with other options. One partial response is a real human result, but it is not the same as a treatment proven to extend life across many patients.

The dual-target, brain-delivered design is a thoughtful answer to two of the biggest reasons CAR-T has struggled against solid tumours: targets that disappear, and cells that never reach the tumour. That's exactly why this trial is being watched. But the honest summary is that it justifies larger, longer trials, not a change in standard care today.

What could this mean if you are considering treatment?

For patients and families facing recurrent glioblastoma, the practical message is twofold. First, CAR-T for brain cancer is a serious, advancing research field — not fringe medicine — and it's reasonable to ask an oncology team whether any clinical trials fit your situation. Second, it remains experimental: outside a trial, this specific therapy isn't an established treatment, and the toxicity profile means it belongs in expert centres with the infrastructure to manage neurological side effects.

This trial can help you ask sharper questions: Is there an open CAR-T trial I might be eligible for? What targets and delivery method does it use? What is the monitoring and recovery like? What happens if neurotoxicity develops? Beware anyone offering "CAR-T for glioblastoma" as a quick, off-the-shelf cure — that is not what the evidence describes.

What we see at the clinic

In Pattaya, families affected by aggressive cancers sometimes ask whether cell therapies like CAR-T are "available" for brain tumours. We try to be straight: this is a fast-moving and genuinely hopeful research area, but for glioblastoma it is still early-phase and trial-based, and the right first step is an honest conversation with a neuro-oncology team about whether a reputable trial is an option. We see a lot of value in helping people tell promising science from premature marketing.

Common questions

Is CAR-T an approved treatment for glioblastoma? No. For glioblastoma it is still experimental and studied in early-phase trials like this one. CAR-T is approved for certain blood cancers, not brain cancer.

Did the therapy cure anyone in this trial? No. It reported one partial response and one case of durable stable disease in a small safety-focused trial ¹ — encouraging early signals, not a cure.

Why deliver the cells into the brain instead of a vein? Because CAR-T cells given into the bloodstream often struggle to reach solid tumours. Infusing them into the brain's fluid puts the therapy closer to the cancer.

[1] Intracerebroventricular bivalent CAR T cells targeting EGFR and IL-13Rα2 in recurrent glioblastoma: a phase 1 trial. Nature Medicine. 2025. https://doi.org/10.1038/s41591-025-03745-0

Key takeaway

This 2025 phase 1 trial showed that dual-target CAR-T cells delivered into the brain are feasible for recurrent glioblastoma, with early signs of benefit in a couple of patients and manageable but real neurological toxicity. It's a meaningful step in one of medicine's hardest problems — best read as careful hope and a reason for larger trials, not as a cure that's ready today.