Bial clinical disaster - a cautious tale

The news that a clinical trial being conducted in France has gone horribly wrong hit the headlines earlier this week.

Currently very little is known about the circumstances which is probably not surprising at this stage. Piecing together various sources, it seems that the drug is called BIA 10-2474 being developed for central nervous system indications such as chronic pain by the Portuguese pharmaceutical company Bial. I have searched for published data on this drug but failed to find anything, which is a little unusual as there’s normally something in the literature by the time of the clinical trials. It’s possible I have missed something but all I could find was the chemical structure and the drug’s target which are given in a US patent application .

BIA 10-2474 is an inhibitor of an enzyme called fatty acid amide hydrolyse (FAAH) which degrades endocannabinoid neurotransmitters. Overall the range of FAAH targets are well established and considered safe (I understand BIA 10-2474 targets FAAH2).  The clinical trial was a Phase-I study, where the drug was given to healthy human volunteers for the first time. It seems that the drug was given as single doses to the volunteers without any issues but problems started to appear when administered as repeat doses.  The media have also reported that the drug was previously tested on animals (which it would have to be under global regulations) but some reports say it was tested in the chimpanzee. 

There are some aspects of the reports so far that look decidedly odd to me. Such a significant switch in toxicity from single doses to repeat doses is extremely unusual and the use of the chimpanzee in toxicology safety testing is very surprising.  Such studies are virtually unheard of these days with the National Institute of Health in the US retiring such studies last year.  It is possible the chimpanzee studies were performed some time ago and it is known that the rodent is not a good model for the FAAH target, but even so this is still very rare.

In the vacuum of information, speculation has started to expand but I think we all have to be cautious, particularly since much of the information we do have comes from the general press, which has a notoriously bad track record when it comes to reporting technical detail. In the early reports the drug was described as cannabis related, presumably because a journalist saw the term “endocannabinoid neurotransmitters” then made an assumption.  Likewise, I wonder if the toxicology species was a monkey (perhaps cynomolgus) which then got translated to chimpanzee - but I could be wrong so this is purely my own speculation.

The incident with BIA 10-2474 has been likened to the previous clinical trial disaster in 2006 with TeGenero’s TGN1412 (TeGenero is now out of business). The two drugs however are very different, with TGN1412 being antibody-based and the resulting toxicity arising from a cytokine storm on the first administration.  BIA 10-2474 is a small molecule drug (molecule weight 327) and apparently only became problematic on repeat dosing. What the two drugs may have in common however, was highlighted in the findings of the TGN1412 enquiry in that animal studies do not always predict adverse effects in humans when the receptor is human specific. There is a smidgeon of this with BIA 10-2474 as the FAAH target is certainly not present in all animal species (and may explain the use of the chimpanzee if reports on the use of this species were accurate). 

There is an argument that the TGN1412 disaster could have been avoided if certain previous studies had been published thus providing a red flag prior to human administration (highlighted in Bad Pharma by Ben Goldacre, pages 8-10). Certainly the industry overall is still plagued with secrecy and remains reluctant to put data into the public domain. There is an ongoing campaign led by Ben Goldacre to make the Pharmaceutical industry publish all human data ( with some companies, such as GSK signing up but the majority still pushing back. It is my view that the regulators should make publication a requirement. Then there would be a level playing field which would remove many of the commercially-based arguments against publishing and the regulators would also be achieving their principal goal of strengthening public safety. Frankly in my opinion, keeping science secret that has such a profound impact on public health is nothing short of criminal. On this basis therefore, it is a little alarming that I can’t find any previously published data on BIA 10-2474.

Having noted the point about transparency in drug testing, we still have to be cautious in the specific case of BIA 10-2474 as we know nothing about the root cause of the disaster. It may, for example, have been a pharmacy dispensing error leading to overdose of the volunteers. (I stress this is pure speculation and there is currently no evidence for this that I am aware of). If it does turn out to be something of this type, then the cause of the BIA 10-2474 disaster would be unrelated to the cause of the one with TGN1412 or the lack of previously published information. We should all keep our powder dry at the moment, I think, on using BIA 10-2474 to promote any specific perceived preventative actions.

In the meantime all we can do is wait for the investigation to report their findings.  I hope the regulatory agencies do not surround themselves in a wall of secrecy and release information as they proceed rather than in a report that might take months or even years to publish. And as we wait, be just a little cautious of what you hear in the general press as it's not always as reliable as we might all want it to be.