The company whose experimental drug left six volunteers critically ill failed to provide regulatory authorities with relevant data before the trial, claim two experts. They say a checklist could help to ensure more rigorous risk assessment for future applications, especially given the increasing popularity of drugs that act on the immune system.

The antibody, called TGN1412, activates the immune system's T cells, and was being developed by the German company TeGenero. But in the first human tests on 13 March, it triggered a dangerous over-reaction of the immune system called a cytokine storm. The agency responsible for approving the trial, the UK Medicines and Healthcare Products Regulatory Agency (MHRA), concluded that this was caused by an unexpected biological effect of the drug.

Now two drug-safety specialists who have investigated the case say relevant information about TGN1412 that might have raised a red flag was missing from the file TeGenero gave the MHRA and other authorities (M. Kenter and A. Cohen Lancet 368, 1387–1391; 2006). “If assessors had known, they probably would have started asking questions,” says Marcel Kenter, executive director of the Central Committee on Research Involving Human Subjects in The Hague, Netherlands, which also assesses drugs for clinical trials. His co-author is Adam Cohen of the Centre for Human Drug Research in Leiden.

Kenter and Cohen identified several key pieces of data that were absent from TeGenero's file. For example, they say that tests in mice given a humanized immune system had shown that the antibody caused T-cell numbers to decline, perhaps because they were so fired up they committed suicide. These data were published online in early March by researchers, including those at TeGenero (N. Legrand et al. Blood 108, 238–245; 2006).

Six volunteers almost died during the TGN1412 trial. Credit: B. STANSALL/AFP/GETTY

The authors also say the company failed to mention key differences between the amino-acid sequence of the antibody receptor in rhesus monkeys on which the drug was initially tested and the human sequence. The differences might have suggested that the antibody would bind more strongly to human than to monkey cells, and perhaps cause a more severe reaction.

Thomas Hanke, chief scientific officer at now-insolvent TeGenero, says all relevant information was given to the MHRA before the trial. Although some of TeGenero's studies were carried out in rhesus monkeys, he says the company used cynomolgus monkeys for toxicological studies because the antibody-receptor sequence of the cynomolgus monkey is identical to the human one.

The MHRA says that at the time of TeGenero's application, senior assessors judged the data submitted to be adequate. But in a preliminary July report into the incident, an expert scientific group recommended that rules for such phase I clinical studies be tightened to improve the safety of volunteers.

The Paul Ehrlich Institute in Langen, Germany, had also approved an application from TeGenero to carry out a similar trial. Its president, Johannes Löwer, says assessors there did ask TeGenero for additional information during the approval process and received it. “I'm not sure [a company] is obliged to include these data but they did provide them, so they were thinking about the their relevance,” he says.

Kenter and Cohen recommend that regulatory bodies such as the MHRA adopt a checklist during risk assessments to ensure that they have all the information on a drug's mechanism of action and the effects of human exposure to similar agents. Löwer agrees, and says the Ehrlich Institute already operates a similar checklist, but that other agencies, with less expertise in basic research, might not.

TGN1412 is thought to be an exceptional case. Even so, Kenter says the lessons are particularly important because of the increasing number of drugs that are designed to work specifically in humans and so may have side effects that are difficult to predict from animal trials.