Drugs trials gone bad, horribly bad

A catastrophic drug safety trial which left six healthy volunteers fighting for their lives was the result of “a powerful pharmacological effect of the product in humans”, the UK government agency charged with investigating the incident said on Wednesday.

The novel drug, called TGN1412, caused multiple organ failure in the six men injected with it at a Northwick Park Hospital, London, UK, on 13 March.

The severe patient reaction in the first human phase I safety trial of the drug was not due to a dosing error, product contamination or manufacturing problems, and the trial was conducted properly according to the agreed protocol, the Medicines and Healthcare products Regulatory Agency (MHRA) said, announcing its interim findings.

The “unprecedented reaction” to the drug was not seen in preclinical test animals at much higher doses, Kent Woods, chief executive officer of the MHRA told reporters. The drug, made by German pharmaceutical company TeGenero, was being tested as a potential treatment for certain autoimmune diseases and leukaemia.

“Even in hindsight, we can see no evidence of any such adverse reactions from the preclinical studies,” Woods said. “The doses given to the humans in the trial were diluted by 500 times – a substantially lower dose than that given to non-human primates, which didn’t show ill effects.”

Five of the six patients have been released from the hospital and the other person is, apparently, doing better. That's good news.

Here are some lessons:

1. Drug trials are inherently dangerous. Don't let a physician tell you that the trial is safe because it's supervised. Yes, the trials are supervised but that doesn't make them safe. I have a friend at NIH (National Institutes of Health in Bethesda, MD, just outside of Washington, DC) and he says trials are better than safe. Why? Because the subjects are constantly monitored and watched by doctors all the time. That is not sufficient to guarantee safety.
   
   
2. Next, lab tests are not enough. There were no problems like these complete organ shut-downs found with lab animals. While animal testing is a necessity, it alone is not enough. Human trials, despite these awful consequences still must be performed. Imagine if the drug were marketed to humans based solely on animal trials? While the problems here are severe, at least they were found before a public release. In that sense, these trials did what they should do: the trials found a fatal problem before the drug could hurt many people.
   
   
3. Computer simulations are bunk. If these trials had such detrimental effects, and they did, and these effects were completely unforeseen, that tells you that computer simulations which are based on what scientists believe they know, are even more likely to fail. Sure, simulations help scientists to think about a problem. They can highlight problems and potential problems. But, even more than animal testing, they are critically limited. If animal tests are not sufficient, simulations are even less sufficient (if that's even possible).
   

What's clear here is that despite the risks, drug trials must be performed on a limited number of humans to proof safety. Humans are too complex, too unpredictable, for scientists to think they predict effects without trials. These trials, while exceptional, saved other people. Thank goodness we still do human testing.