Disastrous TGN1412 Drug Safety Trial In Britain

On March 15th, the first human drug safety trial for new drug TGN1412 went horribly wrong. TGN1412, a monoclonal antibody which stimulates killer T cells, and a potential treatment for rheumatoid arthritis and leukemia, had been tested uneventfully in animals, and it was now time to begin the first human safety tests, on paid healthy young volunteers at doses that were hundreds of times weaker than had been found safe in animals. Shortly after the eight volunteers were given the drug (or, randomly, a placebo), the men started to go down “like dominos.”

Raste Khan, one of the volunteers given a placebo, tells the story in The Times Online:

They took blood samples from everybody. Then they dosed people at two-minute intervals. Roughly five minutes after everyone had been given the drug, the first person who was given the drug started to shake. He took his [shirt] off, he looked like he was burning up, and rubbed his head.

Several minutes later, it missed me and went to the third person. He started doing similar things but he vomited on several occasions. He came back to consciousness but was hyperventilating. He looked like he was in the worst pain.

Then No 4 went through similar symptoms. He took several steps and collapsed. He said, ‘I can’t control myself. I need to use the toilet.’ After that it was like a vomiting bath or something. Nurses had a big black liner for them to vomit in. People were fainting and coming back to consciousness. The gentleman on my left was screaming, saying his back was hurting. It was horrible.

The men were rushed to the hospital, where doctors are attempting to treat them. The Independent reports that one man’s family was told that his head and neck had swollen to three times their normal size, while another’s girlfriend said that he looked “like the Elephant Man.”

Two men are in critical condition, in multiple organ failure, unresponsive, unable to breathe unassisted. Four others are listed in serious but stable condition, though still in intensive care. The remaining two received the placebo, and were unaffected.

New Scientist describes the drug, and speculates as to what may have gone wrong:

TGN1412 is a monoclonal antibody but works slightly differently from other similar drugs. It is a “superagonist”, causing a far greater immune cell response. It also does not require a second, specific trigger to kick-start this response, as do other monoclonal antibodies affecting the same T cell receptor.

“Fortunately, this [super-stimulation] does not occur naturally, because T cells activated in this way would lack any antigenic specificity and could indiscriminately attack normal tissues,” wrote Peter Linsley, from Rosetta Inpharmatics in Seattle, US, in March 2005 in a commentary accompanying a paper in Nature Immunology, which involved TGN1412.

“One could certainly say that, based on what [TeGenero] has already said about TGN1412, the most plausible explanation would be the triggering of a non-specific activation of natural killer T cells leading to indiscriminate cell destruction,” says Ken Campbell, clinical information officer at the Leukaemia Research Fund in London, UK. “This would be consistent with multiple organ failure.”

An immunologist contacted by New Scientist, but who asked not to be named, says: “You don’t need to be a rocket scientist to work out what will happen if you non-specifically activate every T cell in the body.”

Blacktriangle has a rather good round-up of information on the event, with many links to external sources, and points out that Parexel, the U.S. company running the test, may have deviated from the approved testing protocol prepared by the drug’s developer, German company TeGenero:

Mr Khan’s report would seem to be at odds with the statement in this news report that “The protocol indicated that the dosing of subjects was to be staggered over an overall two-hour period.” From his testimony it would appear that everybody would have received the drug (or placebo) within 16 minutes, with the first side effects starting roughly 21 minutes after the start of the process. It will be interesting to see if any investigation concludes that sticking to the “overall two-hour” might have avoided effects in at least some of these individuals.

Authorities took away samples of the drugs that were given to the volunteers, to check for contamination, manufacturing errors, or incorrect dosing. But the incident remains a stark reminder of the limitations of animal models of toxicity in humans: “It’s only a model!”