Michael Eddleston, professor of clinical toxicology at the University of Edinburgh made the suggestion in an editorial in the British Journal of Clinical Pharmacology today in which he argued that preclinical tests that classified BIA 10-2474 as 'low risk' may be lacking.
BIA 10-2474 hit the headlines in January when French regulators halted a Ph I trial after one volunteer died and several others were hospitalised.
In the months since, the activities of Bial and its CRO Biotrial have been under the spotlight, particularly in terms of the design of the Phase I study. The UK Royal Statistical Society suggested there are “clear statistical reservations about the trial’s study design.”
According to a copy of the protocol published by ANSM (after it was leaked by Le Figaro) the study involved 90 volunteers split into several groups, who were given steadily increasing doses of the drug.
Participants were given single doses that gradually increased from 0.25 mg to 100 mg, before being switched to daily doses for 10 days with these doses gradually increasing from 2.5 mg to 50 mg.
No adverse reactions were reported until near the end when a group of six volunteers was due to receive 50 mg for 10 days. One volunteer fell ill after the fifth dose and four others fell ill over the next few days.
The trial design was approved after pharmacological data and preclinical assessments indicated BIA 10-2474 was low risk – potential links to the development of pulmonary lesions in dogs are being investigated. Other FAAH inhibitors have not been associated with adverse events.
For high risk drugs the European Medicines Agency (EMA) advises using sequential dosing, which involves giving the first dose to just one participant and not dosing others until its effects have been assessed.
The Agency made the recommendation after six volunteers were hospitalised during a Phase I trial of TiGenero’s TGN1412.
Sequential dosing may have had a significant impact on the BIA 10-2474 trial according to Eddleston
He told us “If there had been sequential dosing in the multiple ascending dose cohorts, and the delay had been at least until the end of the ten daily doses, then only one person would have been dosed at the toxic dose, and perhaps this person would not have died.”
Eddleston co-authors also called for preclinical data for BIA 10-2474 to be released.
“It is not clear what was reviewed; the summary in the investigators brochure or the full reports. The full reports might have allowed a reviewer to pick up issues that the company had missed and not included in the summary” he said.
Low risk reassessment
But for Eddleston the real learning from the BIA 10-2474 study is that even drugs judged to be “low risk” can be a problem.
He told us that: “Drug pharmacology has to be better described and then studied during the pre-clinical and clinical studies. This aspect will reduce the risk of similar problems occurring - the starting dose for both TGN1412 and this Bial drug appear to have been far too high.”
Source: British Journal of Clinical Pharmacology
“Implications of the BIA-102474-101 study for review of first-into-human clinical trials.”
Michael Eddleston, Adam F Cohen, David J Webb.