Brain Surgery Opens a Short Window for Cancer Drugs in Mice

The surgeon lifts the tumor out, flushes the cavity, checks that the bleeding has stopped, and closes the skull. What’s left behind looks clean. It isn’t. Scattered through the rim of brain around that empty space are glioblastoma cells too invasive to cut and too sparse to see, and clinical follow-up puts the eventual regrowth right there, in the margin, more than 80% of the time.

Standard practice is then to sit tight. Four to six weeks of recovery before radiotherapy and temozolomide chemotherapy begin, by which point the blood-brain barrier, the wall of tight vessel junctions that keeps most drugs out of brain tissue, has sealed itself back up around the survivors.

Work published in Science Translational Medicine suggests that gap is not merely lost time. Researchers at the University of Manchester, with colleagues in Barcelona, resected tumors from the brains of mice and then injected fluorescently tagged liposomes (the fatty nanoscale bubbles that carry doxorubicin in the long-approved formulations Doxil and Caelyx) into the bloodstream at intervals afterward. When the brains were imaged, the pattern turned out to be biphasic. Liposomes given within 15 minutes of surgery piled up at the resection margin, the signal fell away over the following day, and then it surged again in animals dosed 48 to 72 hours after the operation.

Surgery, Not the Tumor, Opens the Gap

The obvious explanation was the cancer itself, since glioblastoma vessels are famously leaky. So the team cut a matching volume of healthy cortex out of mice that had never carried a tumor, and got the same result.

Whatever is opening the barrier in these animals, it seems to be the knife rather than the disease. The window is narrow in space as well as in time: liposomes concentrated within roughly 300 micrometers of the cavity edge, which is precisely where recurrence clusters in patients, and when the injection was held back until a week after surgery the spread had shrunk by 58%. Dynamic contrast-enhanced MRI, the scan clinicians already use to judge barrier integrity, picked up the same leak in the same place, and it had faded to nothing by about eight days.

“For the first time, we’ve shown that glioblastoma surgery briefly exposes a vulnerability we can exploit. If treatment is timed during specific windows we identified, it is able to halt the disease significantly before it regrows,” says Thomas Kisby, principal investigator on the study at Manchester. Everything sitting behind that sentence, for now, happened in mice.

What a Single Dose of Doxorubicin Did in Mice

Doxorubicin is a fine drug against glioblastoma cells in a dish and a useless one in a patient, because it barely crosses an intact barrier and gets pumped back out by the vessel wall when it does. Liposomes, at roughly 80 to 120 nanometers, are far too big to cross at all. Dosed into the open window at 5 mg/kg, though, they ferried the drug straight in. In the first mouse model, a single injection 48 hours after surgery suppressed recurrence for up to 100 days in every treated animal, while the same dose of free doxorubicin did no better than saline. In a second, more infiltrative model, four of 10 treated mice had no detectable tumor at day 80, and the identical dose given eight days after surgery (roughly when a patient might actually receive it in a trial) did almost nothing.

Where the drug ended up is stranger than the outcome suggests. Most of it was taken up by microglia and macrophages, the brain’s resident cleanup crew, rather than by the scattered tumor cells it was aimed at, and the authors say plainly that they do not yet know why the treatment works. A mechanism-shaped hole, right in the middle of the result.

There is reason for caution here, and it isn’t only the usual distance between a mouse and a person. Liposomal doxorubicin has already been through multiple glioblastoma trials in patients, with outcomes ranging from disease stabilization to modest survival benefit, and the argument that bad timing explains those results is a hypothesis this work motivates rather than one it tests; the Manchester press release, for its part, mentions none of the paper’s five stated limitations and widens the story from liposomes, which is what was actually given, to lipid nanoparticle medicines carrying mRNA and genetic payloads, which were not.

Still, the idea is unusually cheap to test. Nothing in it requires a new molecule. The formulation is decades old and already sitting in hospital pharmacies, which is the whole point. “The study has the potential to open a new frontier in postoperative cancer care,” says Kostas Kostarelos, who co-led the work at Manchester and at the Catalan Institute of Nanoscience and Nanotechnology in Barcelona.

Whether the same windows open in a human skull, after a far larger craniotomy and a far messier recovery, is the question the whole proposal rests on, and nobody has looked yet. If they do, the most consequential instrument in the operating room may turn out to be the clock.

Frequently Asked Questions