New insight into BIA 10-2474 clinical trial shines light on preclinical testing

In January 2016, one patient died and four others were hospitalized during a Biotrial-run clinical trial of Bial Laboratory’s molecule, BIA 10-2474.

The molecule was developed to inhibit the activity of fatty acid amide hydrolase (FAAH)​, an enzyme that when inhibited results in the accumulation of fatty acid amides.

In the months following the serious adverse events, much speculation spread as to what went wrong.

As part of its investigation, the French authorities and other experts​ put forth the hypothesis that high doses of BIA 10-2474 inhibited not only FAAH, but other related proteins.

To test this hypothesis, an international group of researchers led by Mario van der Stelt at Leiden Institute of Chemistry and Steven Kushner at the Department of Psychiatry, Erasmus MC, used advanced biochemical methods to define the action profile of BIA 10-2474.

Biotrial, in a comment to Outsourcing-Pharma.com, explained:

“The hypotheses used in this study, carried out by world-renowned researchers for months, are those of an off-target effect of the molecule developed by the Portuguese laboratory Bial​.

“These hypotheses thus confirm once again the unpredictability, for Biotrial, of the accident that occurred during a test which was conducted in accordance with the regulations, and previously authorized by French public authorities​.”

According to the report​ published last week, researchers identified other proteins that were “permanently deactivated following the administration of a high dose of BIA 10-2474​.”

“The Bial compound had many other targets in addition to FAAH and FAAH II​,” Kushner told us.

Tests on human nerve cells confirmed the proteins are involved in the metabolism of fats in nerve cells. Additionally, genetic defects in one of the found proteins had previously been linked to human neurodegenerative disorders, the researchers reported.

“Essentially, all we can say is the drug was not selective compared to other clinically tested FAAH inhibitors and has targets other than the intended target … we cannot say the extent to which, or if at all, the other off targets were the reason people were hurt​,” Kushner explained.

The unanswered questions​

To the extent that is publically known​, Kushner said there is no reason to believe that the animal studies were not done correctly and in accordance with regulatory requirements.

However, the investigation revealed a number of animals “did have some significant problems for which ideally the extent and nature of those problems would have been better described,” ​he explained.

After questions were raised, Bial confirmed​ several dogs in the preclinical trials had been put down as a result of pulmonary lesions.

“That being said, to the extent that we can tell, the regulatory agency went through its procedure; and there is no reason to doubt that procedure​,” said Kushner.

However, the question remains as to what extent the dosing regimen contributed to the complications – a question raised by pharmacology experts following the trial, and one that Kushner said is not for the researchers to answer.

“According to the TSSC [​Temporary Specialist Scientific Committee] established by ​the Director General of ANSM [Agence Nationale de Sécurité du Médicament et des produits de santé],​ one-tenth of the dose used in the trial would have fully engaged the intended FAAH target," ​he explained.

Bial did not respond to request for comment on the report.

Predicting the unpredictable?​

For Kushner, the simple fact is that human biology differs from animals.

“Our brain is dramatically different from that of animals​,” he said, adding that there is reason to believe that the cause of the toxicity may well be specific to humans, or at the very least, distinct from that of the animals used in the testing.

As such, Kushner said activity-based protein profiling (ABPP) of covalent irreversible drug candidates to detect off-targets in human cellular models – such as those used to reveal the off-target proteins of BIA 10-2474 – should be incorporated into drug safety testing.

“If in the drug safety testing procedure, increasingly accurate and realistic human cellular models could be incorporated … it would fill a gap which we think existed here​,” he said.

“Again, we can’t be sure [the off target proteins] were the cause, but had those been identified they certainly would have or should have given the company or regulatory agency some pause and perhaps triggered further investigation​.”

According to Kushner, the fact that the researchers were able to identify the off-target proteins suggests that had ABPP studies of human cells been mandated as part of the preclinical testing procedure of BIA 10-2474, there may have been enough information to trigger a further look into the compound before initiating the first-in-human clinical trial.

Moving forward, Kushner is advocating for increased attention to be paid to advancing the development of better human cell models to detect human-specific toxicity.

“Human cellular testing should be added on top of the current approaches to drug safety testing​,” he said. “There’s no real technical barrier to doing what we did​. This should be within the repertoire of pharma companies that are doing drug development for new compounds​.”

Regulatory agencies have since released statements in support of using human cellular models in preclinical research, though its use remains optional. “The support is there​,” said Kushner. “The question is at what level of urgency​.”