ADePT (NCT07346144)

Glioblastoma is the most common cancer that starts in the adult brain. About eleven thousand people in the United States hear that diagnosis each year, most of them in their sixties. The standard plan is surgery, then radiation with a chemotherapy pill. It helps, but rarely finishes the job. In the trial that set today's standard, half of patients lived under fifteen months.

The reason is the shape of the thing. A glioblastoma has no edge. Its cells wander off into healthy brain, so surgery cannot reach them all. And a few of those cells behave like stem cells. Given time, they rebuild the whole tumor.

So imagine a gene that could kill those cells. Could you put it into a brain safely?

That is the real problem. Delivery is not the hard part; address is. Anything strong enough to kill a tumor cell will damage a neuron if it switches on in the wrong place. Gene therapies handle this with a promoter, a piece of DNA that works as an on switch for a nearby gene. But the switches available are either too weak or too leaky.

So a team in Edinburgh built a better switch.

Start with enhancers. An enhancer is a short piece of DNA that proteins land on to turn a gene up. In nature they cluster into super-enhancers, and those clusters are what give a cell its identity. The team asked whether they could build one from scratch, tuned to glioblastoma.

They began with the two proteins that keep a glioblastoma stem cell in its stem state, SOX two and SOX nine. They chopped the places those proteins bind into short fragments, tested about four and a half thousand of them one at a time, kept the ones that fired only in tumor cells, and chained the winners together. The result is a synthetic super-enhancer. In tumor stem cells it is as strong as the strongest switch in common use. In normal cells it is almost silent.

Then they wired two payloads behind it.

The first is an enzyme borrowed from the cold-sore virus. On its own it does nothing. But it converts a harmless antiviral pill into a drug that poisons dividing cells, so the tumor makes the poison that kills it.

The second is interleukin twelve, an alarm that calls in the immune system. Sent through the bloodstream it is far too toxic. Made inside tumor cells, it stays where it is needed. And because both payloads ride one message, the alarm fades as the tumor cells die.

In mice with an aggressive glioblastoma, one injection cleared the tumor in twenty of twenty-four animals, and they stayed well eleven months. Given fresh tumor cells months later, ten of them grew no tumor at all. Only about a fifth of the tumor cells had to receive the gene. The immune system found the rest.

Mice are not people, and earlier attempts at this kind of therapy struggled in the clinic. That is what this trial is for. It is called ADePT. About sixty-eight adults with new or returning glioblastoma get one injection into the tumor, then the antiviral pill for up to three weeks, then their planned surgery. The first questions are safety and dose. It runs at the Royal Infirmary of Edinburgh with Faye Robertson, and at Ohio State with Bradley Elder.

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