TPT CED rHGG (NCT06666712)

Some brain tumors announce themselves fast. Others take their time. An IDH-mutant glioma is the patient one.

It begins with a typo. IDH is an enzyme in the cell's energy cycle, and in more than seven out of ten lower-grade gliomas, one letter of its gene is wrong. The altered enzyme makes a chemical the body has no use for, and that chemical jams the machinery a cell uses to read its own genes. The tumor that follows grows slowly. It tends to arrive in people in their thirties. And after surgery, radiation and chemotherapy, it comes back. Nearly always, and often more aggressive than before.

When it does come back, there is no standard treatment.

Here is the frustrating part. Drugs that kill these cells in a dish already exist. Topotecan is one. Given by vein, it failed completely in recurrent glioblastoma: in one study, none of twenty-eight patients responded. The drug was not the problem. Getting it to the tumor was.

The brain protects itself. Blood vessels in the brain are sealed shut with tight junctions, and pumps in their walls push foreign molecules back out. It is a good wall, and it cannot tell a poison from a treatment. Raise the dose in the bloodstream enough to force the wall and you poison the rest of the body first.

So what if you skipped the bloodstream?

That is convection-enhanced delivery. A surgeon places a fine catheter into the tumor and the brain around it, and a pump pushes the drug in under gentle pressure. Not diffusion, which fades within millimeters, but bulk flow, which carries a drug centimeters through tissue. Add a contrast agent to the mix and an MRI scan will show you exactly where the drug went.

One more detail shapes the whole design. Topotecan only kills a cell that is actively copying its DNA, and at any moment only a few tumor cells are. A short infusion misses the rest. But the early versions of this ran a tube out through the scalp to a pump at the bedside, and after about four days the risk of infection ends the treatment.

The fix was to put the pump inside the body. A programmable pump, the kind already implanted to deliver pain medicine to the spine, sits under the skin of the chest, is refilled through a needle, and can run as long as it is needed.

In twenty twenty-two, a team at Columbia published the first five patients treated this way. Four infusions of forty-eight hours each, over about a month, with patients at home in between. No severe side effects were blamed on the drug. When the treated tissue was examined afterwards, the share of dividing tumor cells had fallen from roughly four percent to one and a half, and the drug was found as far as six centimeters from the catheter. All five patients eventually died of their disease, and the authors were careful to say that five patients cannot prove benefit.

That study treated a different subtype. This trial brings the method to IDH-mutant tumors for the first time. Six adults whose glioma has returned, at Columbia University in New York, led by Jeffrey Bruce, who has spent a career on this problem. Two operations, four infusions, and one question first. Is it safe?

Everything in this film comes from the published record, and it is not medical advice.