https://orcid.org/0000-0002-6316-3179
Abstract
In 2016, one subject died and four were hospitalized with neurological symptoms during a clinical trial with the fatty acid amide hydrolase (FAAH) inhibitor BIA 10-2474. The present paper reviews the regulatory toxicology studies that were carried out to support the clinical trial application for BIA 10-2474. Animal studies complied with national and international standards including European regulatory guidelines (e.g. EEC Council Directive 75/318/EEC and subsequent amendments). The CNS effects seen in the rat and mouse appear to be common in rodents in such studies and do not in principle seem to be of the type to generate a signal. In the same way in non-human primates, insignificant alterations in the mesencephalon, and especially of the autonomic nervous system (Meissner’s plexus in the bowel) in rodents and monkeys were observed in some animals treated with a high dose. Overall, these data, as well as the extensive additional data generated since the accident, support the conclusion that the tragic fatality that occurred during the clinical trial with BIA 10-2474 was unpredictable and that the mechanism responsible remains unknown, from a non-clinical toxicological perspective.
Introduction
In 2016, one subject died and four were hospitalized with neurological symptoms during a clinical trial with the fatty acid amide hydrolase (FAAH) inhibitor BIA 10-2474. The subject who died, after receiving 50 mg for 5 days, presented evidence of severe brain microhemorrhages and several of the surviving subjects (who received 50 mg for 6 days) also showed evidence of mild to moderate microhemorrhage (Kerbrat et al. Citation2016). The investigations by the French authorities concluded that it was an unexpected effect of the test item, having ruled out other extraneous causes. Their conclusion was that the accident was likely to have been caused by an unknown off-target effect of BIA 10-2474 (CSST Citation2016). Despite several studies and the identification of some off-target interactions with other serine hydrolase enzymes (van Esbroeck et al. Citation2017; Bonifacio, Moser, et al. Citation2018; Bonifacio et al. Citation2020), the mechanism responsible for the toxicity remains unproven.
The clinical trial with BIA 10-2474 was a first-in-man Phase I trial with single-ascending (SAD) and multiple-ascending dose (MAD) phases. During the SAD phase doses of 100 mg demonstrated no safety concerns and during the MAD phase, 10 days of 20 mg were similarly well tolerated (Rocha et al. Citation2016). Inhibition of FAAH increases the levels of several fatty acid amides including anandamide, the endogenous ligand of cannabinoid receptors. Although the primary therapeutic target was analgesia, FAAH inhibitors have been suggested for multiple therapeutic uses and several have been tested in the clinic, with no signs of adverse effects (Li et al. Citation2012; Pawsey et al. Citation2016; Postnov et al. Citation2018). Although directly acting cannabinoid agonists are associated with side effects such as hypothermia, sedation, and impaired memory and attention (De Vry et al. Citation2004; Toczek and Malinowska Citation2018) it is expected that indirectly increasing cannabinoid tone by inhibiting FAAH will be devoid of these side-effects.
During the preclinical development of BIA 10-2474, a comprehensive series of regulatory toxicology studies were performed to support the application for clinical studies. These have been published in detail (Hardisty et al. Citation2020; Harris et al. Citation2020; Hayes Citation2020; Hayes, Hardisty, Harris, Okazaki, et al. Citation2020; Hayes, Hardisty, Harris, Weber Citation2020; Hayes, Pressman, Hardisty, et al. 2020; Hayes, Pressman, Moser, et al. Citation2020; Weber et al. Citation2020). The purpose of this review is to provide an overview of those studies and to explore similarities and differences between the species tested (mouse, rat, dog, and primate) in terms of test item exposure and clinical and pathological signs. Comparisons will be made in four areas: clinical signs, gross pathology, histopathology, and test item pharmacokinetics. Studies included are the 4- and 13-week studies in mouse, rat, dog, and primate, and a 26-week study in rats. Dose-range finding studies will not be discussed in detail, except where they support observations in the other studies. The results of the individual studies will not be presented in detail as they are already available in published reports (Hardisty et al. Citation2020; Harris et al. Citation2020; Hayes Citation2020; Hayes, Hardisty, Harris, Okazaki, et al. Citation2020; Hayes, Hardisty, Harris, Weber Citation2020; Hayes, Pressman, Hardisty, et al. 2020; Hayes, Pressman, Moser, et al. Citation2020; Weber et al. Citation2020). Instead, the data from the different studies and across species will be collated to explore effects that are either common across species or specific to one. The observed effects will be placed in the context of the known pharmacology of BIA 10-2474 and related to the levels of exposure to the test substance.
Clinical signs
Maximum tolerated dose (MTD) studies and up-titration periods
A high dose of 1000 mg/kg was started in mice but this was reduced following unscheduled deaths and signs comparable to those reported below at 500 mg/kg. In the MTD and dose-range finding studies, rats had to be euthanized after 200 or 250 mg/kg/day (p.o.). Clinical signs in these animals included stiff or abnormal gait and/or dragging of fore- and hind-limbs, as well as additional clinical signs such as ruffled fur, salivation, ataxia, and visible weight loss. These signs were also seen in the repeat-dose studies, as discussed below.
In both dogs and primates an up-titration phase was used prior to the fixed-dose phases. This was considered essential to allow the animals to habituate to the acute pharmacological effects of the test substance. In general, the signs observed during the up-titration periods were also seen during the repeat-dose phases. In dogs, after an extended rising-dose phase over 28 days and a repeat-dose phase, one female dog was found dead after 150 mg/kg/day and two other females required euthanasia after 100 and 125 mg/kg/day. Prior to death in these animals, they showed clear signs of sickness: decreased activity and feed intake, hunched posture, decreased grooming (ruffled fur), and labored breathing. There were also signs of ataxia and dragging of limbs.
In the primate, there were some deaths during the MTD studies. Thus, subjects were euthanized after the fourth administration of either 125 or 250 mg/kg/day. In addition, one female was euthanized after the third administration at 60 mg/kg although the other subjects (2 males and one female) completed the study which reached doses of 110 mg/kg/day for 14 days (Weber et al. Citation2020). During the up-titration phase of the 4-week study one female was found dead after receiving four administrations each of 10, 25, and 50 mg/kg/day and nine at 75 mg/kg/day. This subject had shown signs of incoordination, prostration, and weakness. At lower doses, signs including tremors, weakness, and incoordination were seen with increasing frequency from 25 mg/kg/day and from 50 mg/kg/day hypothermia was also seen.
Fixed-dose repeat-administration studies
Only at the highest doses tested during 14-day dose-range finding studies in rats and mice was lethality seen (at 200 and 600 mg/kg, respectively, ). In the 4-week study, at 500 mg/kg, almost a third of the mice died, or required euthanasia, before the end of the scheduled treatment period. At lower doses, and in all other species tested, there were no deaths that were attributed to the test substance in any of the 4- to 26-week studies.
Table 1. Clinical findings observed in toxicology studies in laboratory species.
At doses below those found to be acutely toxic, many of the signs were seen in more than one species, particularly ataxia and weakness, but also decreased activity and tremor which were seen in two species (). If we also take into account 14-day DRF studies autonomic signs included salivation, vomiting, miosis, and hypothermia (although hyperthermia was seen in the dog 4-week study). Most of these signs were only seen at the higher doses tested (500 mg/kg/day in the mouse, 150–200 mg/kg/day in the rat (salivation was seen in the 14-day toxicology study in the rat; Hayes, Hardisty, Harris, Okazaki, Weber Citation2020), 100–125 mg/kg/day in the dog and 50–80 mg/kg/day in the primate). There was a slight tendency for these effects to occur at lower doses as the treatment duration was increased, particularly in the dog for salivation and vomiting but this was far from systematic and largely followed the changes in exposure with chronic treatment (). Weakness, decreased activity, ataxia, and hunched or prostrate posture were seen in all species tested, primarily at the higher doses tested.
Table 2. Hematology findings observed in toxicology studies in laboratory species.
All of the species tested in these studies showed signs of gastrointestinal (GI) tract disturbance. In addition to the vomiting mentioned above, there were pasty or liquid feces in both the dog and primate. There were also widely observed decreases in feed intake and body weight gain which may or may not be related to the GI disturbances. It is also noteworthy that in most cases, these effects tended to occur at lower doses than those inducing most of the other signs. For example, in the primate, vomiting, salivation, and pasty/liquid feces were seen from 6.25 mg/kg/day whereas most other effects occurred at 37.5 and 75 mg/kg/day. Given that there are high levels of cannabinoid receptors and the importance of cannabinoid signaling in the GI system (Hasenoehrl et al. Citation2016; Lee et al. Citation2016), it seems likely that these effects are related to the primary pharmacodynamic effect of BIA 10-2474 to inhibit FAAH and raise levels of fatty acid amides. The effects seen on feed intake may also be an example of exaggerated primary pharmacology as well, as endocannabinoids have a major influence on food intake and energy balance (Borrelli and Izzo Citation2009). In terms of clinical effects, these are the only changes seen at exposure levels corresponding to those seen in human studies (Rocha et al. Citation2016).
Clinical biochemistry and hematology
A broad range of effects on hematology parameters were seen, particularly in the mouse and rat studies, but very few were consistent among species, between studies in the same species, or between males and females in the same study (). Decreases in red blood cell count and hemoglobin were seen in mice (females) and dogs (males, at 20 mg/kg/day in 13-week study only). Changes in blood ion concentrations were only really seen in rats but they occurred at low doses and included increases in sodium, potassium, chloride, calcium, and phosphorus.
There were also increases in cholesterol (rats and mice) and phospholipids, both seen at low doses from 10 mg/kg. It is possible that these changes are related to the reported interaction of BIA 10-2474 with lipid processing enzymes (van Esbroeck et al. Citation2017; Bonifacio, Loureiro, et al. Citation2018; Bonifacio, Moser, et al. Citation2018; Bonifacio et al. Citation2020).
Gross pathology
The most common findings, seen in three of the four species used for toxicology studies, were increases in liver and kidney weights and decreases in thymus size. Increased liver weight was seen in all the rodent studies and in the 4-week dog study (). However, this was not seen in the longer 13-week dog study (which used the same doses as the 4-week study), nor in the primate. Increased kidney weight was seen in rats, dogs and primates but not in the longer 13-week dog and primate studies (). Changes in thymus size were seen in both dog studies in both sexes whereas in the primate this was only seen in males in the shorter 4-week study and in the mouse only in females at the highest dose used in the 13-week study (). The next most common findings were increased lung weights (mouse and dog) and increased adrenal weights (dog and primate). Changes in spleen weight were inconsistent, with decreased spleen weight seen in male rats and male and female dogs whereas, in contrast, increased spleen weights were seen in male mice and male primates in the longer 13-week studies. The increased spleen weights were seen at relatively low doses (75 mg/kg/day in the mouse). It is likely that the enlarged spleens are related to the hematologic effects and related to increased extramedullary hemopoiesis discussed above (Chapman and Azevedo Citation2019). In contrast, decreased spleen weights were seen in male rats in the 26-week study from 30 mg/kg but there were no histological correlates of this. Spleen weights in female rats were unaffected.
Table 3. Necropsy findings observed in toxicology studies in laboratory species.
Histopathology
A summary of the histopathological observations is shown in . Although the mouse showed numerous histopathological effects at the high dose of 500 mg/kg/day, the rat appeared to be the most sensitive species with a range of effects consistently shown at doses of 30 or 90 mg/kg/day across all three studies at 4, 13, and 26 weeks. Many of these signs were also seen in other species, although less consistently than in the rat. Signs of axonal dystrophy or distress, mainly in the medulla oblongata, were seen in all species except the dog and were manifested in almost all cases as vacuolation or swelling. Edema was seen in numerous ganglia of the autonomic nervous system in both the rat and the primate. There seemed to be a particularly high incidence of vacuolar change in ganglia of the GI tract in the rat from 30 mg/kg/day but from 90 mg/kg/day, it was also seen in the ganglia of the salivary gland, prostate gland, and uterus as well as the parathyroid gland. Vacuolation was also seen extensively in Meissner’s plexus throughout the GI tract of the primate from 50 mg/kg/day in the 4-week study and 75 mg/kg/day in the 13-week study. Severity scores in the primate were generally lower than in the rat.
Table 4. Histopathology findings observed in toxicology studies in laboratory species.
In addition to these peripheral nervous system effects, both the rat and the primate showed effects in the central nervous system. In both species, these were seen in the medulla oblongata and consisted of axonal swelling and dystrophy. The rat also displayed axonal swelling in the spinal cord and vacuolation in the ventral commissure of the 3rd ventricle. Some sciatic nerve degeneration was seen in the mouse at high doses but this was not observed in any of the other species tested.
Vacuolation could be related to some of the off-targets previously identified. BIA 10-2474 interacts with the organophosphate target PNPLA6 (also known as neuropathy target esterase) and long-term exposure to organophosphates has been reported to cause vacuolation in sensory ganglia neurons (Rogers-Cotrone et al. Citation2010) as has knockout of the PNPLA6 gene (Akassoglou et al. Citation2004). Although BIA 10-2474 did not affect neurons in these toxicology studies, the behavioral signs (weakness and ataxia) and the histopathological findings (autonomous ganglia vacuolation) are similar to those of unknown etiology reported by Kortz et al in dogs (Kortz et al. Citation1997).
Signs of testicular degeneration were seen in all species except primates, primarily at the highest doses tested.
Reversibility of pathology
Reversibility studies were carried out for the 13-week mouse study, 13- and 26-week studies in the rat, 13-week dog study, and the 4- and 13-week studies in the primate. These were 4-week recovery periods in the mouse and rat, 6-weeks in the dog, and 2 and 4 weeks in the primate for the 4- and 13-week studies, respectively.
In the mouse, many of the effects seen at the end of the treatment phase were absent after the 4-week recovery period, suggesting reversibility. Some were still present such as raised cholesterol levels and increased liver to body weight ratios which, although less elevated than at the end of the treatment period, continued to be higher than in controls. Myofiber degeneration and nephropathy also appeared to be at least partially reversible. In contrast, the sciatic nerve degeneration showed no signs of reversibility.
In the rat, the persistent effects after the recovery period were similar between the 13- and 26-week studies. Thus, the raised plasma cholesterol, phospholipids, triglycerides, and phosphorus were still present after the highest dose at the end of the recovery period and remained outside the historical control range. In the 26-week study in the rat, these measures were often about 30% higher than control values. Other changes remained significantly different from controls but the differences were only greater than ±5% for glucose which was reduced by 8.5% after recovery at 90 mg/kg/day. Although much of the vacuolation that was seen at the end of treatment was absent in the recovery animals, the axonal swelling seen in the medulla oblongata and spinal cord after 30 and 90 mg/kg/day was present at a similar incidence to that noted at the end of treatment. The nephropathy observed after the treatment period was also observed after the recovery period so it did not appear to be reversible.
In the dog, some aphonia was still present in the 100 mg/kg/day group during the recovery period. The macroscopic (discolorations) and microscopic (bronchopneumonia, presence of hemosiderin/hematoidin) changes seen in the lungs were still present after the recovery period.
In the primate 4- and 13-week study all signs were reversible, although some of the clinical signs continued to be observed for 1 or 2 weeks after the end of the treatment period in the shorter study, particularly weakness and tremor.
Reproductive and genotoxicity studies
A series of regulatory studies were carried out to investigate the effects of the FAAH inhibitor BIA 10-2474 on fertility, embryo-fetal toxicity, and pre- and post-natal development in rats and rabbits (Harris et al. Citation2020). Despite some reductions in sperm count in rats from 50 mg/kg/day, there were no major changes in male fertility up to 100 mg/kg/day. In female rats administered up to gestational day (GD) 6, there were reductions in pre-implantation loss at 50 and 100 mg/kg/day but neither post-implantation loss nor early embryonic development was affected. In contrast, when administered to female rats during pregnancy (GD6-GD17), BIA 10-2474 at 75 mg/kg/day reduced feed consumption resulting in weight loss, increased post-implantation loss, and reduced mean fetal body weight. In rabbits, the same maternal toxicity was seen but there were no effects in this species on post-implantation loss or fetal body weights. There were no teratological effects clearly due to BIA 10-2474 and developmental milestones and behavior of offspring were not affected. When administered during pregnancy and lactation (GD6 – post-natal day 20), some post-implantation loss was seen from 20 mg/kg/day, but developmental milestones and behavior of the offspring were not affected, although males tended to have lower body weight.
The genotoxicity of BIA 10-2474 was evaluated using the Ames reverse mutation test (up to 5000 µg/plate), the Escherichia coli WP2uvrA test (up to 5000 µg/plate), an in vitro chromosome damage assay in human lymphocytes (up to 300 µg/ml), and an in vivo micronucleus test in mice (1400 mg/kg/day in females, 1800 mg/kg/day in males) (Hayes, Hardisty, Harris, Weber Citation2020). All results were negative.
Pharmacokinetics
BIA 10-2474 exposure was broadly dose-proportional across all species tested, except for the primate ( and ). In the primate, exposure was consistently less than dose-proportional with exposure increasing by approximately half that expected from the dose increase, particularly at the higher doses. There was some deviation from proportionality after first dose administration, such as in the mouse and dog 4-week studies but this was not seen at the end of the treatment period nor, to any great extent, in other studies in the same species ().
Figure 1. Dose levels and pharmacokinetic parameters (Cmax and AUC0–t) observed in toxicology studies in laboratory species and the clinical trial BIA-102474-101 (single ascending dose [SAD] part; multiple ascending dose [MAD] part).
![Figure 1. Dose levels and pharmacokinetic parameters (Cmax and AUC0–t) observed in toxicology studies in laboratory species and the clinical trial BIA-102474-101 (single ascending dose [SAD] part; multiple ascending dose [MAD] part).](/cms/asset/e7dbdd41-abc7-4a0d-b3ea-e46ae2528296/itxc_a_1867821_f0001_c.jpg)
Table 5. Dose levels (mg/kg/day, p.o.), dose at no-observed-adverse-effect level (NOAEL), corresponding human equivalent dose (HED) and pharmacokinetic parameters (Cmax and AUC0–t) observed in toxicology studies in laboratory species.
Table 6. Summary of male vs female and acute vs chronic differences and dose proportionality for AUC values in BIA 10-2474 toxicology studies.
There was no consistent sex difference except for the rat where females were consistently exposed more than males and repeat treatment saw a slight tendency to increased exposure, although this was less evident in the longer duration studies ().
When corrected for body mass using the FDA guidance values for allometric conversion (FDA Citation2005) and converting to the human equivalent dose (HED), exposure for all species followed a broadly linear extension of that seen in man (Rocha et al. Citation2016) except for the dog where it was lower. Although the data shown are for male animals (), to allow comparison with the male subjects enrolled in the clinical trial, the data for female subjects followed a similar pattern.
It is noteworthy that at all doses used for the toxicology studies, the exposure was higher than that seen in man during the repeat dose clinical phase and after the first administration of the 50 mg dose. In man the mean AUC0–t was 7768 ng.h/mL (with a maximum individual value around 10,000 ng.h/mL) ( and ). The lowest HED calculated from the doses used in animal studies corresponds to 100 mg, which was the highest dose used in the single ascending dose phase in humans and for which no adverse effects were reported ().
Table 7. Dose levels, no-observed-adverse-effect level (NOAEL) and pharmacokinetic parameters (Cmax and AUC0–t) observed in toxicology studies in laboratory species and the clinical trial BIA-102474-101 (single ascending dose [SAD] part; multiple ascending dose [MAD] part).
BIA 10-2474 is metabolized by several routes that are broadly similar across species (rat, dog, and NHP), resulting in the identification of several metabolites in vivo (Loureiro et al. Citation2017). Para- or meta-hydroxylation of the cyclohexyl ring results in BIA 10-2639 and BIA 10-3827 respectively whereas reduction of the pyridine oxide results in BIA 10-2445 which can be further para-hydroxylated to form BIA 10-2631. A final identified metabolite can be formed by demethylation, to form BIA 10-2583.
In the rat brain, BIA 10-2445 was the most prevalent metabolite after administration of 10 mg/kg PO reaching about 20% of the parent Cmax. BIA 10-2631 and BIA 10-2639 were also detected but at less than 10% of parent (Bonifacio et al. Citation2020). Following 5 days of 10 mg/kg PO the detected levels were broadly similar to those seen after acute administration, with no evidence of accumulation of any of the measured substances (Bonifacio et al. Citation2020). In rat plasma, levels of BIA 10-2445 and BIA 10-2583 were analyzed after 1, 97, and 85 days of the 26-week toxicology study at 10, 30, or 90 mg/kg PO of BIA 10-2474, producing AUC levels which did not exceed 3% and 1% respectively of the parent (Hayes, Hardisty, Harris, Okazaki, Weber Citation2020). In the NHP, BIA 10-2445 exposure represented no more than 2.5% of the parent at any of the measured time points in the 4- and 13-week toxicology studies (Weber et al. Citation2020).
In the human clinical trial, BIA 10-2445, BIA 10-2583, BIA 10-2631, and BIA 10-2639 were quantified (Rocha et al. Citation2016). Following acute administration of 40 or 100 mg PO of BIA 10-2474, BIA 10-2583 was not detected, and the other measured metabolites did not exceed 1.5% of the parent for either Cmax or AUC. In the repeat dose phase of the clinical trial, administration of 20 mg of BIA 10-2474 for 10 days resulted in only BIA 10-2639 being detected after the first dose (1.4 and 0.9% of parent Cmax and AUC respectively). After 10 days BIA 10-2639 levels were similar in relation to BIA 10-2474 but BIA 10-2445 and BIA 10-2631 were now detected at between 1 and 4.5% of parent exposure.
Thus, broadly speaking, plasma exposure to BIA 10-2445 (the para-hydroxylated metabolite) was comparable across all species measured, which included rat, NHP, and human. BIA 10-2583 exposure is lower in rats and was absent in humans. After repeat dosing in humans, BIA 10-2631 (the para-hydroxylated and reduced metabolite) became the predominant metabolite but was still at relatively low levels compared to the parent. Although we do not have plasma levels of BIA 10-2631 in any of the toxicology studies, the levels of this metabolite detected in rat brain suggest that, at least in this species, exposure is likely to be similar or exceed that seen in humans.
Pharmacology and off-targets
The pharmacology of BIA 10-2474 has been extensively studied against both human and animal targets (van Esbroeck et al. Citation2017; Bonifacio, Moser, et al. Citation2018; Bonifacio et al. Citation2020). These studies have demonstrated that BIA 10-2474 is a potent inhibitor of FAAH in situ and in vivo. Thus, in the mouse, complete inhibition of FAAH in both liver and brain is obtained at doses of 0.1 mg/kg PO and this inhibition remains above 75% in the brain after 24 h (Bonifacio, Loureiro, et al. Citation2018). At similar concentrations and doses that inhibit FAAH, BIA 10-2474 also inhibits the serine hydrolases alpha, beta hydrolase domain-6 (ABHD6), and carboxyl esterases (van Esbroeck et al. Citation2017; Bonifacio, Moser, et al. Citation2018). Ex vivo studies in rats have demonstrated that as the dose and duration of treatment increases, other serine hydrolases are inhibited. These include lysosomal thioesterase at 30 mg/kg after 28 days and, after 100 mg/kg for at least 7 days, ABHD11, group XV phospholipase A2, PNPLA6, ABHD10, and AIG1 protein (Bonifacio, Moser, et al. Citation2018; Bonifacio et al. Citation2020). Importantly, using human cell lines in situ, van Esbroeck et al. (van Esbroeck et al. Citation2017) identified a very similar set of serine hydrolase enzymes inhibited by BIA 10-2474 but they required concentrations of 10–50 µM applied over 24 h. The overall similarity of the off-targets identified in human cells in situ and in the rat in vivo study after 28 days (where activity of metabolites was assessed) supports the conclusion that none of the metabolites formed in the rat are present in sufficient concentration to affect additional off-targets other than those reported for the parent BIA 10-2474.
Conclusions
The clinical signs reported by Kerbrat et al. (Kerbrat et al. Citation2016) included headache, a cerebellar syndrome, memory impairment, and altered consciousness. Magnetic resonance imaging showed bilateral and symmetric cerebral lesions, including microhemorrhages predominantly involving the pons and hippocampi. These signs were seen after 4 or 5 administrations of 50 mg. There is no record of such signs being seen at the lower dose of 20 mg after 10 days administration, nor after the single administration of 100 mg during the single ascending dose phase (Gama et al. Citation2016). The AUC and Cmax values associated with this 100 mg dose (23,166 ng.h/mL and 1772 ng/mL, respectively) are comparable to the values obtained at the lowest doses in the rat and the non-human primate ( and ). Although values are not available for the repeat dose phase at 50 mg, the data following the first administration for this dose and the repeat-dose data for 20 mg suggest that at 50 mg, the corresponding values are likely to be approximately half those attained at the single dose of 100 mg (extrapolation from data presented by Rocha et al. Citation2016). It is therefore very likely that the exposure to BIA 10-2474 at the lowest doses used in the toxicity studies exceeded that seen in the clinic. This conclusion probably also applies to the metabolites of BIA 10-2474 although the absence of clinical data on the metabolites present in the clinical cohorts makes this suggestion necessarily tentative (Huang et al. Citation2019).
Table 8. Dose levels, abnormal findings and pharmacokinetic parameters (Cmax and AUC0–t) observed in toxicology studies in laboratory species and the clinical trial (single ascending dose [SAD] part; multiple ascending dose [MAD] part).
Although many of the clinical signs were not (or could not) be assessed during the preclinical toxicity studies, the cerebellar syndrome described in the human subjects (which included limb, gait, and postural ataxia) did have parallels in the animal studies. Gait (except in monkeys) and postural ataxia (rodents only) were seen in tested species. Similarly, although limited changes were seen in the hippocampus (only seen in two mice after 500 mg/kg/day, and in one rat after 150 mg/kg/day), signs of brainstem degeneration were seen in the medulla oblongata of the rat after 13 weeks at 30 mg/kg/day and in the non-human primate after 4 weeks at 50 mg/kg/day or 13 weeks at 75 mg/kg/day. At these doses, the plasma exposure is nearly an order of magnitude greater than that measured at the toxic dose in the clinic ().
The NOAEL doses established by the CROs running these studies and accepted as valid by the independent commission set up by the French authorities (Agence Nationale de Sécurité du Medicament) are shown in . The exposure (Cmax and AUC) shift margins over NOAELs calculated from these doses are all greater than the dose responsible for the fatal event in the clinical trial ().
At the doses used in the toxicity studies (apart from the lowest dose in some studies), all of the off-targets identified so far would have been occupied to some extent (van Esbroeck et al. Citation2017; Bonifacio, Moser, et al. Citation2018; Huang et al. Citation2019; Bonifacio et al. Citation2020). In contrast, at the observed clinical exposure in the multiple dose phase, it is unlikely that they would be occupied based on the in situ data presented by van Esbroeck et al. (Citation2017) using human cells. Furthermore, there is no evidence for accumulation of either BIA 10-2474, or its metabolites, either in plasma (all species described here) or in several brain regions of the rat (Loureiro et al. Citation2017; Bonifacio, Loureiro, et al. Citation2018; Bonifacio et al. Citation2020). Thus, although signs of cerebellar disturbance seen in humans did have similarities with signs seen in some animal studies (ataxia and gait disturbances), the exposure required was much higher. This is supported by the observation that in animal studies other signs were seen at the same, or lower, dose (). In particular, none of the more serious toxicity seen in humans was seen in animals and it is noteworthy in this respect that the committee set up by the French authorities to examine the clinical trial accident concluded that no toxicity comparable to that observed in the clinical trial accident was demonstrated in any of the animal studies (CSST Citation2016). Thus, the microhemorrhages and stroke-like effects were not observed in any of the animal studies. The absence of this specific toxicity in animals is also supported by the absence of any facilitation of cerebral stroke or stroke-like pathology in an animal model susceptible to hemorrhagic stroke (Bonifacio et al. Citation2020).
In conclusion, there is nothing in the toxicology studies that predicts the type of pathology (stroke) seen in the clinical trial. Although there were some superficially similar observations in some animal studies, it is not clear that these are related to the effects seen in the clinic. Indeed, the CSST report (CSST Citation2016) suggested that the CNS effects seen in the rat “appear to be common in rodents in such studies and does not in principle seem to be of the type to generate a signal. In the same way in non-human primates and rats, cerebral damage and especially of the autonomic nervous system (Meissner’s plexus in the bowel) was observed in some animals treated with a high dose.” Overall, these data, as well as the extensive additional data generated since the accident, support the conclusion that the tragic fatality that occurred during the clinical trial with BIA 10-2474 remains unexplained and none of the data available to Bial and the authorities prior to the clinical trial, nor any of the extensive data collected by BIAL or others since (van Esbroeck et al. Citation2017; Bonifacio et al. Citation2020), would have allowed us to predict this toxicity. Furthermore, the mechanism responsible for the toxicity remains unknown.
| Abbreviations | ||
| FAAH | = | Fatty acid amide hydrolase |
| MTD | = | Maximum tolerated dose |
| DRF | = | Dose range finding |
| NHP | = | Non-human primate |
| p.o. | = | Oral |
| GI | = | Gastrointestinal |
| PNPLA6 | = | Neuropathy target esterase |
| Cmax | = | Maximum concentration |
| AUC | = | Area under the curve |
| HED | = | Human Equivalent Dose |
| NOAEL | = | No Observed Adverse Effect |
Acknowledgments
The manuscript was written by the authors and reviewed by Bial-Portela & Ca S.A. (São Mamede do Coronado, Portugal); however, the work product and conclusions are those of the authors. The authors thank the editor and his reviewers (Who remain unknown to us.) for their most helpful comments and suggestions that greatly improved the scientific basis and readability of the manuscript.
Declaration of interest
These studies were funded by Bial-Portela & Ca S.A. (São Mamede do Coronado, Portugal). A. Wallace Hayes and Klaus Weber were paid consultants of Bial-Portela & Ca. S. A. Paul Moser and Patrício Soares-da-Silva are employees of Bial Portela & Companhia. S. A. BIA 10-2474 has been the subject of a Phase 1 clinical trial in France (Rennes) which was interrupted because of the occurrence of adverse reactions and is currently the subject of pending proceedings in France.
References
- Akassoglou K, Malester B, Xu J, Tessarollo L, Rosenbluth J, Chao MV. 2004. Brain-specific deletion of neuropathy target esterase/swisscheese results in neurodegeneration. Proc Natl Acad Sci U S A. 101:5075–5080.
- Bonifacio MJ, Loureiro A, Aires C, Fernandes-Lopes C, Sousa F, Palma N, Moser P, Soares-da-Silva P. 2018. Pharmacodynamic and ADME evaluation of FAAH inhibitor BIA 10-2474. Mol Biol Cell. 28:630. Abstract P2651.
- Bonifacio MJ, Moser P, Soares-da-Silva P. 2018. In-vivo selectivity profiling of BIA 10-2474 against serine hydrolases in the rat brain. Mol Biol Cell. 28:629. Abstract P2650.
- Bonifacio MJ, Sousa F, Aires C, Loureiro AI, Fernandes-Lopes C, Pires NM, Palma PN, Moser P, Soares-da-Silva P. 2020. Preclinical pharmacological evaluation of the fatty acid amide hydrolase inhibitor BIA 10-2474. Br J Pharmacol. 177:2123–2142.
- Borrelli F, Izzo AA. 2009. Role of acylethanolamides in the gastrointestinal tract with special reference to food intake and energy balance. Best Pract Res Clin Endocrinol Metab. 23:33–49.
- Chapman J, Azevedo AM. 2019. Splenomegaly. StatPearls. [updated 2019 May 6].
- De Vry J, Jentzsch KR, Kuhl E, Eckel G. 2004. Behavioral effects of cannabinoids show differential sensitivity to cannabinoid receptor blockade and tolerance development. Behav Pharmacol. 15:1–12.
- FDA. 2005. Estimating the maximum safe starting dose in initial clinical trials for therapeutics in adult healthy volunteers.
- Gama H, Vieira M, Graça J, Chassard D, Massano J, Patat A, Santos A, Rocha F, Soares-da-Silva P. 2016. Preliminary safety evaluation in study BIA-10-2474-101. Proc British Pharmacol Soc. 16(1):188. http://wwwpa2onlineorg/abstract/abstractjsp?abid=33277&kw=2474&author=Rocha&cat=-1&period=-1
- Hardisty JF, Harris SB, Hayes WA, Weber K. 2020. Oral repeated-dose toxicity studies of BIA 10-2474 in beagle dogs. Regul Toxicol Pharmacol. 111:104555.
- Harris SB, Hardisty JF, Hayes AW, Weber K. 2020. Developmental and reproductive toxicity studies of BIA 10-2474. Regul Toxicol Pharmacol. 111:104543.
- Hasenoehrl C, Taschler U, Storr M, Schicho R. 2016. The gastrointestinal tract - a central organ of cannabinoid signaling in health and disease. Neurogastroenterol Motil. 28:1765–1780.
- Hayes AW. 2020. Commentary on BIA 10-2474. Regul Toxicol Pharmacol. 111:104541.
- Hayes AW, Hardisty JF, Harris SB, Okazaki Y, Weber K. 2020. Oral repeated-dose toxicity studies of BIA 10-2474 in Wistar rat. Regul Toxicol Pharmacol. 111:104540.
- Hayes AW, Hardisty JF, Harris SB, Weber K. 2020. The absence of genotoxicity of a novel fatty acid amide hydrolase inhibitor, BIA 10-2474. Regul Toxicol Pharmacol. 111:104556.
- Hayes AW, Pressman P, Hardisty JF, Harris SB, Weber K. 2020. Oral repeated-dose toxicity studies of BIA 10-2474 in CD-1 mice. Regul Toxicol Pharmacol. 111:104557.
- Hayes AW, Pressman P, Moser P, Soares-da-Silva P. 2020. Regulatory safety pharmacology evaluation of BIA 10-2474. J Pharmacol Toxicol Methods. 102:106677.
- Huang Z, Ogasawara D, Seneviratne UI, Cognetta AB 3rd, Am Ende CW, Nason DM, Lapham K, Litchfield J, Johnson DS, Cravatt BF. 2019. Global portrait of protein targets of metabolites of the neurotoxic compound BIA 10-2474. ACS Chem Biol. 14:192–197.
- Kerbrat A, Ferre JC, Fillatre P, Ronziere T, Vannier S, Carsin-Nicol B, Lavoue S, Verin M, Gauvrit JY, Le Tulzo Y, et al. 2016. Acute Neurologic Disorder from an Inhibitor of Fatty Acid Amide Hydrolase. N Engl J Med. 375:1717–1725.
- Kortz GD, Meier WA, Higgins RJ, French RA, McKiernan BC, Fatzer R, Zachary JF. 1997. Neuronal vacuolation and spinocerebellar degeneration in young Rottweiler dogs. Vet Pathol. 34:296–302.
- Lee Y, Jo J, Chung HY, Pothoulakis C, Im E. 2016. Endocannabinoids in the gastrointestinal tract. Am J Physiol Gastrointest Liver Physiol. 311:G655–G666.
- Li GL, Winter H, Arends R, Jay GW, Le V, Young T, Huggins JP. 2012. Assessment of the pharmacology and tolerability of PF-04457845, an irreversible inhibitor of fatty acid amide hydrolase-1, in healthy subjects. Br J Clin Pharmacol. 73:706–716.
- Loureiro AI, Lopes C, Bonifacio MJ, Moser P, Soares-da-Silva P. 2017. Pharmacokinetics and pharmacodynamics of BIA 10-2474, a novel FAAH inhibitor, following oral administration in the cynomolgus monkey. Proc British Pharmacol Soc. 16(1):212. http://wwwpa2onlineorg/abstract/abstractjsp?abid=33301&author=soares-da-silva&cat=-1&period=64
- CSST. 2016. Report by the Temporary Specialist Scientific Committee (TSSC): FAAH (Fatty Acid Amide Hydrolase), on the causes of the accident during a Phase 1 clinical trial in Rennes in January 2016. In: Santé ANdSdMedPd, editor. France: Saint-Denis.
- Pawsey S, Wood M, Browne H, Donaldson K, Christie M, Warrington S. 2016. Safety, tolerability and pharmacokinetics of FAAH inhibitor V158866: a double-blind, randomised, placebo-controlled phase I study in healthy volunteers. Drugs R D. 16:181–191.
- Postnov A, Schmidt ME, Pemberton DJ, de Hoon J, van Hecken A, van den Boer M, Zannikos P, van der Ark P, Palmer JA, Rassnick S, et al. 2018. Fatty acid amide hydrolase inhibition by JNJ-42165279: a multiple-ascending dose and a positron emission tomography study in healthy volunteers. Clin Transl Sci. 11:397–404.
- Rocha F, Santos A, Chassard D, Patat A, Astruc A, Falcão A, Gama H, Soares-da-Silva P. 2016. Tolerability, pharmacokinetic and pharmacodynamic profile of BIA 10-2474 in healthy volunteers following multiple ascending doses. Proc British Pharmacol Soc. 16(1):178. http://wwwpa2onlineorg/abstract/abstractjsp?abid=33267&kw=2474&author=Rocha&cat=-1&period=-1
- Rogers-Cotrone T, Burgess MP, Hancock SH, Hinckley J, Lowe K, Ehrich MF, Jortner BS. 2010. Vacuolation of sensory ganglion neuron cytoplasm in rats with long-term exposure to organophosphates. Toxicol Pathol. 38:554–559.
- Toczek M, Malinowska B. 2018. Enhanced endocannabinoid tone as a potential target of pharmacotherapy. Life Sci. 204:20–45.
- van Esbroeck ACM, Janssen APA, Cognetta AB 3rd, Ogasawara D, Shpak G, van der Kroeg M, Kantae V, Baggelaar MP, de Vrij FMS, Deng H, et al. 2017. Activity-based protein profiling reveals off-target proteins of the FAAH inhibitor BIA 10-2474. Science. 356:1084–1087.
- Weber K, Hacker R, Hardisty JF, Harris SB, Hayes AW. 2020. Oral repeated-dose toxicity studies of BIA 10-2474 in cynomolgus monkeys. Regul Toxicol Pharmacol. 111:104547.