Key information was absent from a research file prior to a London drug trial that left six men seriously ill, investigators have concluded.
All drug trials carry a risk
Drug research experts from the Netherlands say UK regulators did not receive findings that might have warned them of damage TGN1412 could do.
The worst affected of the six men, Ryan Wilson, had fingers and toes amputated, and his future health is uncertain.
In The Lancet, the authors propose a system of extra checks for trial drugs.
An expert group set up by the UK's Health Secretary is looking at how to safeguard against such adverse events when experimental drugs are tested.
A preliminary report by the group recommends the initial drug dose should be given to just one person - which was not the case in the London trial.
And it recommends drugs which affect the immune system, like the monoclonal antibody TGN1412, may be best given to people who are already ill.
The Lancet paper authors tested their own proposed list of safety checks using the case of TGN1412.
The checks include factors such as how much is known about how the experimental drug works in the lab, in animals and in humans and how predictable these effects are.
In the case of TGN1412, the scientists from the German company TeGenero reported that the site in the body where the drug binds was identical in humans beings and monkeys.
However, no detailed data on such a comparison was included.
When the Dutch researchers explored this they found clear differences between humans and monkeys.
The research file on TGN1412 also lacked information about how the drug affects certain human immune cells compared to monkey immune cells.
Dr Adam Cohen, from the Centre for Human Drug Research in Leiden, the Netherlands, said: "Essential information was absent.
"The assessors did not receive all relevant findings."
Dr Cohen and his team say drug developers need to ensure they provide all essential information, and trial assessors need to use internationally consistent ways to evaluate which human experiments should go ahead.
A spokesman for the UK's drug trial assessor, the Medicines and Healthcare products Regulatory Agency, said: "We conducted a rigorous assessment of the application by experienced assessors who concluded it was reasonable to proceed with the trial."
He said the MHRA had implemented recommendations since made by the UK's independent expert group on the way such trials are reviewed in the future.
"Clinical trials in general have an excellent safety record but it is important that we learn the lessons from the TGN 1412 incident," he added.