Clinical Trials in Safety Spotlight

In January of this year, one man lost his life and 5 others were hospitalized after taking the experimental drug Bia 10-2474 during a Phase I clinical trial in Rennes, France. Since the tragedy took place, two in-depth investigations have been carried out: one by the Temporary Specialist Scientific Committee (TSSC) set up by the French medicines agency (1), and the other by the Inspection Générale des Affaires Sociales (IGAS), the inspectorate for social affairs in France (2). In light of the findings, the European Medicines Agency has initiated a review of its guidelines for first-in-human clinical trials – with the aim of better ensuring the safety of human volunteers (3).

The IGAS report concluded that Biotrial (the French contract-research organization who conducted the trial) and Bial (the Portuguese pharmaceutical company who sponsored it) are partly responsible for what happened based on the dose they chose to administer and the time they took to alert the authorities and other trial participants. The TSSC report concluded that the accident was related to the molecule tested, and that Bial and Biotrial both acted within the current rules, but followed a “flawed testing protocol”. The report was very critical of Bial’s Investigator Brochure, which “contains many mistakes, inaccuracies, figure inversions… making understanding difficult in several aspects.” The authors describe their findings as “highly surprising given the regulatory importance of this document.” Though the authors did not comment on whether or not the trial should have been authorized, they pointed out that “of the 63 pages of the Investigator Brochure summarizing the preclinical data, fewer than two discuss demonstration of pharmacological activity for the apparently planned indication.” According to the authors, this meant that it wasn’t possible to determine an effective dose before “never risk‐free” preclinical and clinical development took place. The report also revealed that preclinical tests indicated that BIA 10-2474 had a lower efficacy than the comparator product – a fact that was deleted from the Investigator Brochure. The authors also said they were “astonished” that volunteer selections did not include a neurophysiological assessment.

The EMA says it will take the findings of both reports into consideration when improving best practices and guidance in current protocols – and aims to produce a concept paper by July that identifies areas for change. To start the process, the EMA has established two expert groups to carry out preparatory work. One group will look at pre-clinical aspects and the data needed to safely initiate first tests in humans; the other group will examine clinical aspects and how the design of first-in-human trials could be improved.

Although the French trial has put first-in-human studies in the spotlight, the EMA explains that such studies are usually very safe. Since 2005, approximately 3,100 first-in-human studies have been carried out in Europe and only one other severe incident has ever been reported.