Experts: Drug trial disaster followed several missteps; new rules needed

French Health Minister Marisol Touraine

Why did an experimental "me-too" drug tested in France kill one patient and leave 5 others hospitalized?

French health authorities gathered a committee of experts together to explore that question, and their first stab at an answer leaves the mystery largely unsolved. The committee, though, zeroed in on unexpected off-target effects of the drug as a likely culprit for the serious brain damage inflicted on trial subjects. And there were enough visible mistakes in the study--a poor preclinical evaluation of pharmacology, the use of maximum doses that were far higher than what was needed to achieve the desired effect, and erratic dose escalation in the human study--that the authors are calling for new trial rules that should be adopted globally.

The drug in question--a painkiller dubbed BIA 10‐2474--is an FAAH inhibitor similar to others that have been abandoned; not because they were unsafe, but because they were ineffective. "Its originality is relative as it could appear to be a 'me‐too' of several molecules previously developed as FAAH inhibitors such as PF‐3845 by Pfizer and JNJ‐42165279 by Janssen (J&J)," the scientists report.

The Portuguese pharma company Bial had billed the compound as a reversible FAAH inhibitor, but the drug in fact appears to be irreversible. In addition, by blocking FAAH you increase levels of an endocannabinoid called anandamide, which in turn can spur the development of other compounds "which could have a harmful effect, especially on brain circulation."

"BIA 10‐2474 would appear to be a lot less specific to FAAH than its predecessors," the experts also note, "making binding to other cerebral enzymes plausible."

Particularly perplexing for the experts was safety data from studies using the drug on four different species: rats, mice, dogs and monkeys. Several primates were "put down" in the study, but only after taking high doses of the drug. Cerebral damage was noted in rodents, but only after they were given doses more than 100 times stronger than what was planned for humans. In the dogs, though, there were changes in their lungs that raised questions, along with a query about why investigators in the study lowered the dose as they went along.

None of that, though, would prevent an investigator from trying the drug in humans, the report states. What was particularly troubling, they said, was that the drug didn't show the kind of efficacy as an analgesic that warranted a human study: "This seems too basic to justify continuing development, a fortiori in humans." Better preclinical pharmacology studies should have been required.

The scientists raised questions about some of the patients recruited, including one with a head injury. Dose escalations were sometimes too sharp, and they once again singled out the fact that the investigators continued to dose patients even after one of the patients in the study was hospitalized.

Patients in only one of 14 cohorts in the study experienced severe adverse events; "serious central nervous system symptoms exclusively, only appeared in the exposed volunteers from MAD cohort 5 (50 mg).

So what could have happened? The scientists speculated that there could have been an interaction with other products; "anandamide‐related toxic effects" could have been involved; toxicity from a metabolite of the drug may have played a role.

One favorite theory is that the drug had an off-target effect, hitting other cerebral enzymes, which may have been likely given the random nature of the FAAH inhibition they saw in the drug. And cohort 5 was given doses up to 40 times what was needed to achieve full inhibition of FAAH, raising additional questions about the trial design.

The group also wants new rules put in place that govern trial design, which they're recommending for international use.

First and foremost, they want to see stricter requirements on preclinical pharmacological studies to demonstrate that a drug has potential. Trial volunteers should be screened better for CNS studies, including a neuropsychological assessment. Maximum doses should be better evaluated and new, best practices rules on dose escalation need to be adopted.

- here's the report (PDF)