Cancerous pediatric brain tumors are some of the most aggressive cancers to affect children, and are frequently fatal. They’re difficult to treat due to their proximity to sensitive brain tissue in tiny brains, and children’s bodies can rarely tolerate the side effects of the levels of chemotherapy and radiation necessary to shrink tumors.

But recently, researchers at Stanford Medicine, the Lucile Packard Children’s Hospital, and several other institutions successfully tested a promising immunotherapy treatment that shrank multiple tumor types in mouse models. Immunotherapy treatments harness the body’s own immune system to fight the cancer, and usually come with few to no side effects compared to chemotherapy drugs and radiation.

The collaborative study, published in Science Translational Medicine, showed results on the five most common types of pediatric tumors: Group 3 medulloblastomas (MB), atypical teratoid rhabdoid tumors (ATRT), primitive neuroectodermal tumors (PNET), pediatric glioblastoma (PG), and diffuse intrinsic pontine glioma (DIPG).

The Stanford researchers designed their study after the recent discovery of a molecule known as CD47, a protein expressed on the surface of all cells. CD47 sends a “don’t eat me” signal to the immune system’s macrophages—white blood cells whose job it is to destroy abnormal cells. “Think of the macrophages as the Pac-Man of the immune system,” Samuel Cheshier, lead study author and assistant professor of neurosurgery at Stanford Medicine, tells mental_floss.

Cancer cells have adapted to express high amounts of CD47, essentially fooling the immune system into not destroying their cells, which allows tumors to flourish. Cheshier and his team theorized that if they could block the CD47 signals on cancer cells, the macrophages would identify the cells on the cancerous tumors and eat them—without any toxicity to healthy cells. To do so, they used an antibody known as anti-CD47, which, as its name implies, blocks CD47 on the cancer from binding to a receptor on the macrophage called SIRP-alpha.

“It is this binding that tells the macrophage, 'Don't eat the tumor,'” he says. The anti-CD47 fits perfectly into the binding pocket where CD47 and SIRP-alpha interact, “like a jigsaw puzzle,” helping the macrophage correctly identify the tumor as something to be removed. “Anti-CD47 is the big power pill in Pac-Man that makes him able to eat the ghosts,” says Cheshier.

Even better, not only does anti-CD47 block the “don’t eat me” signal, it has the rare ability to pass the blood brain barrier, making it “very effective against all brain tumor types,” Cheshier says.

His team tested anti-CD47 on each of the five tumor types both in vitro (in live tissue cultured in a dish) and in vivo (human cancer cells implanted inside living mice). For the initial in vitro studies, Cheshier explains, they developed the cancer cells “in a way that preserves the cancer stem cells and allows them to grow.” Then the researchers introduced macrophages and added anti-CD47. Excitingly, the scientists “watched the macrophages eat the tumors,” he says.

Next, for each of the five tumor types, they isolated two separate lines of cancer cells taken from separate patients and cultured all 10 in the lab. Then they injected each of these different lines of tumor cells directly into the brains of 10 to 20 mice, so that a minimum of 20 animals per tumor type were tested. The tumor cells had been modified with firefly luciferase genes, making the tumors light up under scans so the scientists could track the cells’ progress. “Once we confirmed the tumor was growing, we gave some mice anti-CD47,” Cheshier says, while the control mice received none. “Only the mice that received anti-CD47 lived,” he explains.

Additionally, the scientists created an experiment where they also injected healthy human brain stem cells into mouse brains, in addition to tumor cells, and then treated some of the mice with anti-CD47. “In the mice that received anti-CD47, the normal brain cells grew normally. So there was no effect on [the healthy cells] even in the context of very active tumor killing.” Cheshier finds this result exciting because “this is the first time in any study where anyone put normal human cells with the cancer and then showed that anti-CD47 wasn’t toxic in an animal.”

Depending on the tumor type, and the amount of anti-CD47 injected, the tumors visibly shrank over the course of one week to six months—even disappearing altogether in some cases. While not every tumor was completely eliminated, Cheshier says this is most likely a matter of the length of the experiment and the amount of anti-CD47 given. “We could achieve [elimination] in every tumor type,” he says.

Of course, Cheshier warns that humans might react differently than mice, but these initial results are promising: He is most excited that a single therapy worked in all five tumor types. “You can imagine a situation where instead of giving different types of drugs for different tumors, we can just say, ‘Here is the treatment. It’s universal.’”

The pre-clinical study took four years to complete, and now the therapy has moved to phase one of a human clinical trial process, in order to test for toxicity. He has plans for phase two, which will ask the question, “Does it actually work in treating the tumor [in humans]?” Cheshier says. And phase three will be the randomized, double-blind clinical trial that hopefully proves anti-CD47 will be superior to current treatments. Meanwhile, other studies will look at its combined effects with other cancer treatments.

While there’s much further work to be done, Cheshier is very optimistic that this therapy will be both “more effective and less toxic than current standards of care. I think anti-CD47 will be part of an armamentarium where we’re using the immune system to treat cancer instead of toxic chemicals and radiation beams.”