Curing someone of cancer is, in theory, a piece of cake: all you have to do is kill the cancer cells while leaving the healthy cells intact.

But in practice, this solution is far more difficult. In fact, it remains one of the great unsolved problems in modern oncology: how do you find, target and destroy each individual cancer cell in the body—while minimizing damage to the surrounding cells.

But luckily, Harvard Stem Cell Institute scientists at Massachusetts General Hospital may have finally struck gold: they have designed special, toxin-secreting stem cells that can target and destroy brain tumors. Their findings, which were performed in laboratory mice and which appear in the latest issue of the journal STEM CELLS, offer up an entirely unique method for eradicating deadly cancers.

Harvard Neuroscientist Khalid Shah, who led the study, explained in last Friday’s news release that the idea of engineering stem cells to kill cancer cells is not new—but there was a key difference in scientists’ ability to target individual cells vs. difficult-to-reach tumors, which is often the case with brain cancer.

The solution, Shah and his team argued, was stem cells. Previously, Shah and his team discovered that stem cells could be used to circumvent these problems. The fact that stem cells continuously renew meant that they could also be used to continually deliver toxins to brain tumors

In this study, the research team introduced a small genetic change, or mutation, into the stem cells so that they become impervious to the toxin’s harmful effects. They then introduced a second mutation that allowed the stem cells to maintain and produce and secrete toxins throughout the cells’ lifetime—effectively giving it an unlimited supply of ammunition to use once it encountered the brain tumor.

They then employed a common technique whereby the toxins were tagged so that they only sought out and infected cancer cells—leaving healthy cells unscathed.

While preliminary, these results are encouraging. As the team continues to refine their method of development and delivery, they are optimistic that they can bring their methods to clinical trial within the next five years.