[UPDATE]: The Drug Safety French agency has released the protocol used by Bial and approved by the agency (PDF in French), according to an article published in Nature (http://www.nature.com/news/researchers-question-design-of-fatal-french-clinical-trial-1.19221?WT.mc_id=FBK_NatureNews): Press Release and PDF file in French
If you have been following the news, you may have heard about the tragic ending of a Phase I clinical trial carried in Rennes (France) for a new painkiller that was firstly described as a cannabis-like compound before being refuted.
The drug was set for a Phase I trial, with an initial 90 volunteers, with repeated injections of a same dose, in a placebo-controlled experimental setup. Out of 90 volunteers, 4 of them are in serious clinical conditions, one was declared brain dead within 24 hours and and as I write this post, his death was just announced in one French newspaper (http://www.dna.fr/actualite/2016/01/17/essai-therapeutique-le-patient-en-etat-de-mort-cerebrale-decede)
After leaving a few days to have the fog of breaking news and conflicting reports dissipate, I thought it would be helpful to write down a post to summarize the information about the drug candidate, the operating procedure behind Phase I trial and the current conditions on human experimentation. This is not an exhaustive review and comments are welcome to help improve this article.
What was the drug candidate tested?
According to the article posted in the French daily newspaper “Dernieres Nouvelles d’Alsace” dated from January 17th 2016 (see link above), the drug candidate that was used in the clinical trial is “BIA-10-2474”. For any chemist enthusiast around, here is the chemical structure of BIA, according to Pierre van de Weghe (Universite Rennes 1, Rennes, France).
BIA-10-2474 (that I will summarize as BIA) in the rest of the text is an experimental inhibitor of the fatty acid amid hydrolase (FAAH), an enzyme involved in the anandamide, an endogenous ligand of endocannabinoid system (as depicted in the following diagram).
The endocannabinoid system is the same system targeted by 9-THC one the active compound found in cannabis. Aside from the hallucinogenic effects, it also has some analgesic properties. Therefore, by acting on the endocannabinoid system, that may represent an alternative to opiates.
Bial, a Portugese pharmaceutical company has developed BIA as one of their product that was in their pipeline as depicted by the company website:
BIA was not the only drug candidate in the pipeline, as various other compounds especially targeting neurological disorders such as epilepsy and Parkinson’s disease went through Phase III clinical trials and are few steps away from approval by EU drug approval authorities.
BIA was a complete internal product, with no peer-reviewed publications listed in Pubmed (this is fairly common, due to important risk of being scooped by other pharmaceutical companies capable to synthesize such compounds through reverse engineering).
What is a Phase I clinical trial and what happened?
Drug discovery in pharmaceutical research follows the same pattern. Basic sciences identify some novel mechanism of a disease. Because you identify a novel signaling pathway driven by a protein, a nucleic acid, a sugar or a fatty acid derivative.
Pharmaceutical companies have huge chemical libraries either synthesized by medicinal chemists or extracted from natural compounds, that can contains up to millions of different compounds. Once you have identified a novel target for a drug and developed a biological assay, these companies will screen their whole libraries in a procedure called “high throughput screening” (HTS) run by automated robots capable to runs thousands of compounds per day.
Once they find a positive result, a “hit”, then they identify the compound and by fine medicinal chemistry, works on refining the biological activity by synthesizing a lead compound with hundreds of similar differing by subtle changes in the structure.
Once they found the best-matching compound, it undergoes screening using cell-based assays (in vitro) and animal-based assays (in vivo) to demonstrate if it has any potential. If it has any therapeutical potential, it goes into pharmacokinetics and drug metabolism testing to determine how this compound is handled by the whole organism as well as determining the toxicology (to determine any toxic effect, in particular any neurotoxicity, cardiotoxicity, hepatotoxicity and renal toxicity). This step is important at it prelude the Phase I clinical trial. Once it crosses these different checkpoint, the lucky one gets into Phase I trial. In Phase I clinical trial, it is all about assessing the compound is safe and determine any toxicity or side effects, but also determine the maximum dose that is deemed safe (this is what we call the minimum toxic concentration or MTC). Most of the time, the phase I is run in healthy volunteers, but sometimes it may be performed in patients volunteers that are affected by a condition deemed incurable in their late stage of the disease.
In the case of BIA, it was done on healthy volunteers, it was done through a contractor named “Biotrial” under the authorization by the French Agency of Drug Safety as described in their website (http://www.ansm.sante.fr/S-informer/Actualite/La-survenue-d-effets-graves-ayant-entraine-l-hospitalisation-de-6-patients-dont-un-en-etat-de-mort-cerebrale-a-conduit-a-l-arret-premature-d-un-essai-clinique-du-laboratoire-BIAL-Point-d-information).
Biomedical experiments in human subject is a sensitive issue that have seen various situations in humankind. Until the Second World War, experimentation on human subjects were fairly unregulated and looked more a “wild west” in which physicians and scientists had a large degree of maneuver, sometimes at the expense of the volunteer integrity. However, that radically changed how human subjects were involved in experiments. During the second World War, nazi scientists (with some Nazi physicians such as “Dr. Mengele” became celebrity due to their most outrageous “medical” experiments) found in concentration camps a trove of human subjects to perform their experiments that not only had little or no scientific merit by their experimental design or rational at best, the worst form of torture or sadism that one human being can apply to another.
After the end of the Second World War, an international court named as the “Nuremberg Trial” took place in Nuremberg, Germany in 1946 to prosecute the SS chain of command that lead to one of the worst form of genocide documented by the mankind. After such trial, a second one took place and was aimed at prosecuting the Nazi physicians. This second trial served to set fundamental basis in the terms of involvement of human subjects in medical experiments named as the “Nuremberg Code” (https://en.wikipedia.org/wiki/Nuremberg_Code).
A key aspect of the Nuremberg Code is the absence of financial compensation that was deemed to avoid the pressure of human subjects to enroll and remain in the experiment.
Financial compensation is not obligatory but is seldom as a compensation for the time volunteers have allocated themselves for the experiment. In that case, an article from Science (http://www.sciencemag.org/news/2016/01/more-details-emerge-fateful-french-drug-trial) mentioned a honorarium of EUR1900 in addition of travel fare reimbursement. The experimental design was done for a 2-weeks period, with 10 oral administration and at least 40 blood sampling (likely for the PK and toxicology study).
The first round of the trial was done successfully in the single dose setting, with no adverse effect noted. However, the situation turned when the third stage of the trial designed to find the highest dose and the effect of repeated dose. This stage enrolled 90 volunteers, males and females in a age range of 18-55. Five of them presented same symptoms, with one that showed cerebral death and died as I wrote this post.
According the same article from Science, they cite the Chief of the Neurosciences Pole of Rennes University Hospital as following “… neurologist Gilles Edan of the University of Rennes Hospital Center said yesterday. MRI imaging has shown “deep, necrotic and hemorrhagic lesions in the brain” of the patients, Edan said.“. Similar information was also reported by Forbes (http://www.forbes.com/sites/judystone/2016/01/16/bials-french-clinicial-trial-ends-in-disaster-what-went-wrong/#2715e4857a0b309e43439b2c), but the Forbes article is raising rightfully some interesting questions as quoted in their article “There are so many questions, as little information has been released. We don’t know what doses this group of ill volunteers received, or how that was different from earlier groups. It was likely the first of a higher dosage. We don’t know if or how food affected the drug’s metabolism. Could there have been a contaminant causing disaster in one batch of drug? While much less likely with an oral than IV drug, an error in manufacturing or the concentration of drug could perhaps have affected one batch and not others.”
There are still a lot of information that remains unknown.
One speculation I may raise based on the allegation is the report of “necrosis and brain hemorrhage” that would raise for me the possible disruption of the blood-brain barrier in these patients, leading to a massive brain hemorrhage and the subsequent brain damage. Such logical explanation is further supported by the facts the physicians have extensively used MRI scans to assess the damage.
Science article on this story mentioned that other companies like Pfizer found no undesirable effects of their home-made FAAH inhibitor but dropped the compound after failing to find clinical efficacy in Phase II clinical trial. The impact of FAAH and the barrier function remains unclear but there is a study by Piomelli and colleagues (Moreno-Sanz G. et al, Pharm Res 2014) that mentioned another FAAH inhibitor that share some chemical structural similarities with BIA suggested the interactions with P-gp (ABCB1) and BCRP (ABCG2), two drug efflux pump at the BBB (http://www.ncbi.nlm.nih.gov/pubmed/24993496). Another study done by O’Sullivan and colleagues (Hind WH et al., Br J Pharmacol 2015) highlighted the protective effect of anandamide and oleoylethanolamide (two endogenous cannabinoid ligands and FAAH substrate) as protecting the barrier function at the BBB using an in vitro model of stroke injury (http://www.ncbi.nlm.nih.gov/pubmed/25651941).
Hopefully, we can hope to learn about what went wrong in the next few weeks.