In 2006 Ryan Wilson, a healthy 20 year old Londoner, volunteered for a first-in-human clinical trial to help test the safety of a new drug, TGN1412, intended to treat rheumatoid arthritis and leukemia. The cash he’d get in exchange for his time would help fund his upcoming vacation.

Instead, he nearly died.

Even though the drug amount injected in his body was 500 times lower than the dose found to be safe in animals, Wilson experienced a catastrophic immune reaction, called a cytokine storm, that led to heart, kidney and liver failure, pneumonia and the loss of his toes and three fingers to dry gangrene. The other five healthy volunteers were also severely injured.

TGN1412’s devastating effect was unfortunately missed in preclinical laboratory and animal studies prior to the human trial. Unlike the pills in your medicine cabinet which are made up of synthesized chemicals, TGN1424 belongs to a growing class of medicines called biologics which come from biological sources such as proteins, DNA, sugars and cells. There is a concern that once a biologic is injected in a patient, the immune system may mount a strong attack all over the body. If that happens, too many immune cells, or white blood cells, are activated and release proteins, called cytokines, which in turn activate more immune cells and the reaction spirals into a dangerous cytokine storm like in Ryan Wilson’s case.

Clearly this tragedy begs for tests that can better predict drug toxicity in humans well before the first trial participants step into the clinic. On Monday a research team from the Imperial College London reported in the journal FASEB that they have done just that using human blood stem cells.

The team’s novel test is not so different than previous ones. Both tests are carried out in a petri dish using two human cell types: white blood cells and endothelial cells, a component of blood vessels. Both tests are also designed to mimic the human immune system’s response to biologics by measuring the release of cytokines.

But the Imperial College London team’s test differs from others in one important way: both the white blood cell and endothelial cell types come from the same individual. First they collect a donor’s blood stem cells and specialize them into endothelial cells. Then white blood cells are also collected from the same donor.

The prior tests, on the other hand, rely on cells from two different donors. Because the two cell types aren’t necessarily tissue-matched, the white blood cells may already be primed for an immune response even before a biologic is added to the test. In fact, these prior tests weren’t able to distinguish between a biologic known to cause a limited immune response versus TGN1424, known to cause a cytokine storm. The newly developed test, however, accurately predicts both the toxic cytokine storm caused by TGN1424 and the absence of a response by several approved biologics, such as the breast cancer drug Herceptin.

In a college news release, Jane Mitchell, the senior author on the report, sees the big picture importance of her lab’s work:

“As biological therapies become more mainstream, it’s more likely that drugs being tested on humans for the first time will have unexpected and potentially catastrophic effects. We’ve used adult stem cell technology to develop a laboratory test that could prevent another disaster like the TGN1412 trial.”

Their results also highlight the often-overlooked power of stem cells to not just deliver therapies but to help develop safer ones.