The Pathway to Proof‐of‐Concept for BNC210, a Negative Allosteric Modulator of the Alpha‐7 Nicotinic Acetylcholine Receptor (nAChR), for Treatment of Psychiatric Disease

Abstract

BNC210 is an investigational small molecule selective negative allosteric modulator of the alpha‐7 nicotinic acetylcholine receptor (α7 nAChR). It is an anxiolytic compound with a novel mechanism of action. In a series of Phase 1 clinical trials in healthy volunteers, psychometric test batteries showed that BNC210 did not cause attention, cognition, or memory impairment, negative effects on mood or emotional stability, sedation, or addiction, ruling out undesirable side effects of known anxiolytic compounds. In healthy volunteers, target engagement at the α7 nAChR was demonstrated in a nicotine shift assay using quantitative electroencephalography, and BNC210 demonstrated improvement in panic‐like symptoms in a cholecystokinin tetrapeptide panic model. Initial clinical trials used an aqueous suspension formulation of BNC210 to cover a wide dosage range; however, its pharmacokinetic parameters were consistent with solubility‐limited absorption and a significant food effect. A dispersible tablet formulation was then developed with improved bioavailability and is being used in Phases 2 and 3 clinical trials. Collectively, the Phase 1 data demonstrated desired properties of BNC210 supporting proof‐of‐concept clinical trials. BNC210 is currently being developed for acute, as‐needed treatment of social anxiety disorder and chronic treatment of post‐traumatic stress disorder.

BNC210 is an investigational small molecule selective negative allosteric modulator of the alpha‐7 nicotinic acetylcholine receptor (α7 nAChR) ligand‐gated cation channel. In animal studies, the α7 subtype nAChR modifies anxiolytic behavior, ¹ and antagonism at the nAChR in the amygdala reduces fear‐related behavior. ² Thus, modulation of the α7 nAChR may be a beneficial target for the treatment of anxiety and stress‐related disorders. Preclinical pharmacology studies with BNC210 demonstrated that it was effective in several mouse and rat behavioral models of anxiety and stress‐related disorders over a broad therapeutic dosage range. ³

In Phase 2 clinical trials, BNC210 demonstrated efficacy in patients with generalized anxiety disorder (GAD), social anxiety disorder, and post‐traumatic stress disorder (PTSD). In a single dose Phase 2 trial in patients with GAD, BNC210 produced a significant reduction in anxiety‐induced brain abnormalities, measured using functional magnetic resonance imaging, including bilateral amygdala reaction to fearful faces in the Emotional Faces Task and connectivity between the left amygdala and anterior cingulate cortex. ⁴ , ⁵ BNC210 also significantly reduced the intensity of threat avoidance behavior and self‐reported anxiety. ⁶ Another Phase 2 trial evaluated as‐needed acute dosing of BNC210 in social anxiety disorder patients (NCT05193409) using a simulated public speaking challenge and found improvements in anxiety levels in the participants on BNC210 compared to placebo. ⁷ In a 12‐week Phase 2 clinical trial of patients with PTSD (NCT02933606), BNC210 was effective at significantly reducing PTSD total symptom severity scores, depression, and insomnia compared to placebo. ⁸ Results from these clinical trials support further clinical evaluation of BNC210 for treating anxiety and stressor‐related disorders.

Prior to the Phase 2 trials, Phase 1 trials were conducted to characterize the safety, tolerability, and pharmacokinetics (PK) of BNC210. BNC210 is an anxiolytic compound with a novel mechanism of action. Two potential safety issues needed to be investigated based on the side effect profiles of known anxiolytic compounds, such as benzodiazepines, and the pro‐cognitive properties of nicotine and other α7 nAChR agonists raising the question of whether a negative allosteric modulator of α7 nAChR would have any detrimental effects on cognition. In preclinical pharmacology and toxicology studies BNC210 did not elicit sedation, memory or motor impairment, or physical dependence when evaluated in a number of rodent models, thus lacking many of the undesirable side effects associated with known anxiolytic drugs, and it showed no signs of cognitive impairment. ³ Here we report on the lack of side effects and lack of cognitive impairment of BNC210 using psychometric test batteries in a series of Phase 1 trials in healthy volunteers.

The rationale for developing a negative allosteric modulator is that there may be significant inhibition of the α7 pathway with a modulator under high states of arousal, such as in anxiety and stress‐related disorders, whereas under basal states, there would be less inhibition. This could lead to state‐dependent pharmacological activity, which would improve the therapeutic index over traditional anxiolytics, which can be accompanied by high burdens of sedative adverse effects due to their tonic action. However, demonstrating this activity in humans presents a challenge, especially without a suitable positron emission tomography (PET) ligand.

We report on the results of a quantitative electroencephalogram (qEEG) paradigm with a nicotine shift assay used to demonstrate target engagement in healthy human volunteers. Since BNC210 is a negative allosteric modulator of the α7 nicotinic receptor, treatment with BNC210 was postulated to modify the response of the α2 band on the qEEG recordings to nicotine. Results from a pharmacological model of panic in a Phase 1 trial, along with the other Phase 1 data, provided confidence to proceed with proof‐of‐concept Phase 2 clinical trials with BNC210 in anxiety and stressor‐related disorders.

Pharmacokinetics of an early aqueous suspension formulation were evaluated in healthy volunteers, and this formulation was successfully used for the Phase 2 single‐dose clinical trial in GAD patients. However, when used in a 12‐week treatment out‐patient Phase 2 trial in PTSD patients instructed to take BNC210 with food (NCT02933606), a population PK analysis showed that plasma levels were appreciably lower than expected and were highly variable. This finding indicated that the suspension formulation was unsuitable for administration outside of a controlled clinic setting. An exposure‐response model established a concentration‐effect relationship indicating the potential for BNC210 to have benefit in PTSD, but because of the poor performance of the suspension formulation, an overall effect by intention‐to‐treat was not achieved in this trial. ⁹ A spray‐dried dispersion tablet formulation was then developed to overcome the solubility and bioavailability limitations of the aqueous suspension and was used for the Phase 2 social anxiety disorder trial (NCT05193409) and the second trial in PTSD patients (NCT02933606), both of which demonstrated preliminary clinical benefit of BNC210.

Methods

Key design characteristics of the clinical trials described in this paper are listed in Table 1, and sampling schedules for key pharmacokinetic and pharmacodynamic assessments are listed in Table S1. All trials were conducted in healthy volunteers and all recruited male participants only, except trial BNC210.011 which enrolled an equal number of male and female participants. Participants were aged 18 to 35 (BNC210.004), 45 (BNC210.003), 55 (BNC210.005) or 65 (all other trials) years of age, were in good general health without clinically significant renal, hepatic, cardiac or respiratory disease, had no significant history of illicit substance abuse or excessive alcohol use, and were required to have washed out from all prescription and over‐the‐counter medications prior to trial entry. All trials were approved by a local Human Research Ethics Committee (HREC)/Institutional Review Board (IRB), and all participants gave written informed consent. Details of the trial sites and applicable HRECs/IRBs are provided in Table S2.

Table 1.

Summary of BNC210 Clinical Trials in Healthy Volunteers

Trial identifier	Trial design	Participants enrolled (No. Receiving BNC210)	Doses (mg)	Oral formulation
BNC210.001	SAD, R, DB, PC, PK, PD, tolerability	32 (24)	5, 15, 50, 150, 300, 600, 1200, and 2000 (single dose; fasted)	Aqueous suspension
BNC210.002	SAD, R, DB, PC, PK (food effect), PD, tolerability	4 (3)	300 (single dose; fasted), 300, 600, 1200, and 2000 (single dose; fed)	Aqueous suspension
BNC210.003	4‐way CO, R, DB, DD, PC, lorazepam‐controlled, PD	24 (22)	300 and 2000 (single dose; fed)	Aqueous suspension
BNC210.004	2‐way CO, R, DB, PC, PD	60 (59)	2000 (single dose; fed)	Aqueous suspension
BNC210.005	MAD, R, DB, PC, PK, PD, and tolerability	56 (44)	150, 300, 600, and 1000 mg BID for 8 days (fed)	Aqueous suspension
BNC210.009	3‐way CO, OL, PK, tolerability	6 (6)	300 (single dose; fasted or fed)	Aqueous suspension and tablet (150 mg)
BNC210.010	SAD, OL, PK, tolerability	5 (5)	600, 900, and 1200 (single dose; fasted)	Tablet (150 mg)
BNC210.011	PK, OL, tolerability	10 (10)	900 (single dose, Days 1 and 7), 900 (BID, Days 2 to 6) (fed)	Tablet (225 mg)

BID, twice daily (bis in die); CO, cross‐over; DB, double‐blind; DD, double‐dummy; MAD, multiple ascending dose; OL, open‐label; PC, placebo‐controlled; PD, pharmacodynamics; PK, pharmacokinetics; R, randomized; SAD, single ascending dose.

Formulations

Three formulations were evaluated in the Phase 1 trials: an aqueous suspension formulation composed of BNC210 in hydroxypropyl methylcellulose (HPMC, 0.5% w/v), solutol (0.1% w/v), benzyl alcohol (0.5% v/v), and water, and two dose strength tablet formulations (150 and 225 mg) containing a spray‐dried dispersion of either 30:70 (w/w) BNC210:HPMCAS‐MG (HPMC acetate succinate, grade MG) or 35:65 (w/w) BNC210:HPMCAS‐MG, respectively.

Doses and Dose Administration

In the initial clinical trials, an aqueous suspension formulation was used. This was for simplicity of manufacture and the ability to cover a wide dosage range. These trials included:

• A placebo‐controlled first‐in‐human single ascending dose (SAD) trial (BNC210.001) with eight sequential cohorts, each of four participants (three on BNC210 and one on placebo). The BNC210 doses ranged from 5 to 2000 mg administered after an overnight fast.

• A placebo‐controlled SAD trial (BNC210.002) to determine whether the plateau in absorption seen in the fasted state in trial BNC210.001 could be overcome by dosing BNC210 with food, which involved one cohort of participants (three on BNC210 and one on placebo). Single BNC210 doses of 300 mg (in both fasted and fed states) and 600, 1200, and 2000 mg (in the fed state) were administered with a minimum 7‐day wash‐out between doses. In the fed state, participants consumed an FDA‐standard high‐fat breakfast. ¹⁰

• A 4‐way cross‐over, double‐dummy, placebo‐ and lorazepam‐controlled trial (BNC210.003) in 24 participants who received single doses of 300 and 2000 mg BNC210 (fed state, high‐fat) with a lorazepam placebo, 2 mg lorazepam with a BNC210 placebo, and a double placebo, with a minimum 5‐day wash‐out between doses. Since the predicted times to reach peak plasma concentrations (Tmax) are different for lorazepam and BNC210, lorazepam was administered 3.5 h after BNC210 in a blinded double‐dummy manner.

• A 2‐way cross‐over, placebo‐controlled trial which included an injection of cholecystokinin tetrapeptide (CCK‐4) (BNC210.004) in 59 participants who received single doses of 2000 mg BNC210 (fed state, high‐fat) and placebo with a minimum 5‐day wash‐out between doses.

• A placebo‐controlled multiple ascending dose (MAD) trial (BNC210.005) with three sequential cohorts each of eight participants (six on BNC210 and two on placebo) evaluating BNC210 doses of 150, 300, and 600 mg administered twice daily (BID; 300, 600, and 1200 mg/day), and a fourth cohort of 30 participants (24 on BNC210 and six on placebo) at the 1000 mg BID dose level (2000 mg/day). All participants were dosed for 8 days and received their trial medication twice a day after finishing a standard meal.

A spray‐dried dispersion tablet formulation was later developed to overcome the bioavailability limitations of the aqueous suspension. Three open‐label clinical trials evaluated the PK of the tablet formulation:

• A 3‐way cross‐over, single‐dose trial (BNC210.009) to compare the PK of the suspension and a prototype 150 mg‐strength tablet. One cohort of six participants all received 300 mg BNC210 as the suspension (fasted) and the tablet formulation [in both the fasted and fed (high‐fat) states] with a minimum 5‐day wash‐out between doses.

• A SAD trial (BNC210.010) with one cohort of five participants who all received sequential doses of 600, 900, and 1200 mg BNC210 using the 150 mg‐strength tablet in the fasted state with a minimum 5‐day wash‐out between doses.

• A 7‐day trial (BNC210.011) of the 225 mg‐strength tablet in one cohort of ten participants (five female and five male) who received 900 mg of BNC210 as a single dose on Days 1 and 7, and BID on Days 2 to 6 (1800 mg/day), after consuming a standard meal.

Pharmacokinetic Evaluations

Blood sampling for plasma PK analyses was carried out in trials BNC210.001, BNC210.002, BNC210.005, BNC210.009, BNC210.010, and BNC210.011 according to the schedules in Table S1. Details of the bioanalytical methods used for PK analyses are in the Supplementary Information. PK analysis was by the standard non compartmental method.

Pharmacodynamic Evaluations

In the first 2 trials with BNC210 (BNC210.001 and BNC210.002), a 16‐item Bond and Lader visual analogue scale (VAS) ¹¹ was used to determine whether there was a particular profile of central nervous system (CNS)‐attributable symptoms associated with BNC210 administration, as there was no relevant human precedent on which to base likely desired and undesired effects.

Following that, a comprehensive examination of potential CNS effects was conducted in trial BNC210.003, where lorazepam, a benzodiazepine, was also included as a positive control. The psychometric test battery included:

• 1.
  Measures known to be sensitive to sedating drugs:

  • a.
    Multiple Choice Reaction Time ¹² is an evaluation of attention and vigilance that measures the time needed to have a motor response to a visual stimulus
  • b.
    Critical Flicker Fusion Threshold ¹³ is an evaluation of attention and vigilance that measures visual processing speed
  • c.
    Digit Symbol Substitution Test ¹⁴ assesses speed of information processing and sustained attention by measuring the number of correct answers on a test
  • d.
    Karolinska Sleepiness Scale ¹⁵ is a 9‐point self‐rated scale that measures subjective feelings of sleepiness
  • e.
    Peak Saccadic Velocity ¹⁶ is an objective recording of eye movement velocity, which measures visuo‐motor coordination
  • f.
    Perceptual Priming Test ¹⁷ is used to measure an unconscious, non‐intentional form of memory, called implicit memory
• 2.
  Measures of CNS activity:

  • a.
    Emotional‐VAS (eVAS) ¹⁸ is a self‐rated 10 cm‐line scale with a sad face at one end and a happy face at the other end, used for evaluation of emotional state
  • b.
    Quantitative electroencephalogram (qEEG) is used for psychopharmacological investigation of new psychotropic compounds. ¹⁹
• 3.
  A measure of addiction liability:

  • a.
    Addiction Research Center Inventory (ARCI 49) ²⁰ is a 49‐item test that measures the broad range of physical, emotive, cognitive, and subjective effects of drugs, including (1) amphetamine‐like effects (e.g., increased energy, sense of well‐being), (2) benzedrine‐like effects (e.g., increased energy, intellectual productivity), (3) morphine‐benzedrine‐like effects (e.g., pleasant somatic experiences, euphoria), (4) lysergic acid diethylamide‐like effects (e.g., dysphoria, somatic discomfort), and (5) pentobarbital‐chlorpromazine‐alcohol‐like effects (e.g., sedation, psychomotor retardation).
• 4.
  Neuroendocrine biomarkers:

  • a.
    Serum cortisol
  • b.
    Serum adrenocorticotropic hormone (ACTH)

The tests were performed at 3.5, 6, 9, and 12‐h post‐BNC210/placebo dose.

Further cognitive testing was performed in trial BNC210.005 using a proprietary psychometric test battery called Plateforme d’Études PSYchométriques (PEPSY). ²¹ The tests included Choice Reaction Time, Digit Vigilance, Rapid Visual Information Processing, Learning Memory Task, Numeric Working Memory, Spatial Working Memory, Bond and Lader VAS, and ARCI 49. The tests were administered at 6 h post‐dose (around the predicted T_max for BNC210) on Day 1 and Day 8.

Additionally, in trial BNC210.005, evidence of target engagement for BNC210 at the α7 nAChR was sought using a nicotine shift assay. This consisted of a succession of qEEG recordings to record the amplitude of response of the α2 band (10 to 12.5 Hz) following an intranasal titration of nicotine up to a cumulative dose of 2 mg. After 8 min of recordings (i.e., four 2 min‐long qEEG recordings in the following sequence: vigilance controlled, then eyes closed, then vigilance‐controlled, then eyes closed and resting) to establish baselines (only the eyes‐closed epochs were used), a first dose of 0.5 mg nicotine (one puff in one nostril) was administered and the 2‐min qEEG recording started 5 min later. Then, if the nicotine was still well tolerated (based on the absence of nausea, dizziness, and palpitations), another puff was given, alternating nostrils, and recordings were performed up to 2 mg of cumulative nicotine dose (i.e., 4 puffs separated by 7 min each). On Day 7, the nicotine challenge was done 6 h post‐BNC210/placebo dose, and the baseline on Day‐1 was done at approximately the same time of day.

The effect of BNC210 on induced panic‐like symptoms was evaluated following a 50 µg injection of CCK‐4, which is a known anxiety and panic inducer in susceptible people ²² (BNC210.004). The CCK‐4 injection was timed around the expected T_max for BNC210 (7 h post‐dose). Participant responses were measured using the Panic Symptom Scale (PSS), an instrument containing 18 items derived from the diagnostic criteria of a panic attack as defined in the DSM‐III‐R, ²³ and an eVAS.

Results

Pharmacokinetics

Suspension Formulation

After administration in the fasted state, peak plasma levels of BNC210 occurred 1 to 2 h post‐dosing and declined following first order kinetics (BNC210.001; Table S3). Minimal drug was excreted unchanged in the urine across the dose range of 5 to 2000 mg (13.6% to 0.6% of the dose administered, respectively). At dose levels up to 50 mg, area‐under‐the‐concentration‐time curve (AUC) and maximum plasma concentration (C_max) values increased approximately dose proportionately. With further increase in dose, the exposure continued to increase up to 600 mg, but there was no appreciable further increase in exposure beyond that. Given the low aqueous solubility of BNC210, this plateau in absorption with increasing dose was attributed to solubility‐limited absorption, hence, the effect of food on absorption was investigated.

Trial BNC210.002 (Table S4) showed a more than doubling of C_max and an increase in AUC of 3.5‐fold when dosing BNC210 with food at the 300 mg dose. In contrast to the fasted state, exposure continued to increase at 600 and 1200 mg in an approximately dose‐proportional manner, but the increase in AUC at 2000 mg was only 15% greater than that at 1200 mg. Time to maximal concentration (T_max) was delayed to between 5 and 7 h in the fed state.

In the MAD trial with BID dosing (BNC210.005; Table S5), a 1.8‐ to 2.9‐fold increase in C_max and AUC_0‐12 was observed after repeated doses (Day 8) compared to a single dose (Day 1) for all dose groups. Steady state was achieved on Day 2 following multiple administrations of 150 and 300 mg BID, and on Day 6 following multiple administrations of 600 and 1000 mg BID. Overall, increases in exposures were less than dose‐proportional with increasing dose levels, and the increase in steady state AUC_0–12 at 1000 mg BID was only 14% greater than that at 600 mg BID. Mean steady state AUC_0–12 achieved with 1000 mg BID (2000 mg/day) was 24.9 mg h/L.

Tablet Formulation

A 150 mg‐strength prototype tablet formulation using spray‐dried dispersion technology was developed to overcome the solubility and bioavailability limitations of the suspension formulation. This was directly compared with the suspension formulation in a 300 mg single dose PK trial (BNC210.009; Table S6) and produced a marked improvement in absorption in the fasted state compared to the suspension, where C_max and AUC were higher by 2.5‐fold and 2.3‐fold, respectively (Figure 1). Furthermore, there was a much smaller food effect with the tablet, where AUC only increased by 1.3‐fold with food, compared to 3.5‐fold with the suspension in trial BNC210.002. T_max for the tablet was increased in the presence of food from 2 h (fasted) to 4 h (fed). In the SAD trial with the 150 mg‐strength tablet (BNC210.010; Table S7), exposure in terms of both C_max and AUC increased proportionally with dose, in the fasted state, up to the highest dose level of 1200 mg (Figure 1).

Figure 1.

Mean (+SE) BNC210 plasma concentration‐time profiles following single dose administration of the suspension and tablet formulations under fasted and fed conditions (Trials BNC210.009 and BNC210.010).

The tablet formulation was further optimized, and in the multiple dose trial with the 225 mg‐strength tablet administered over a 7‐day period at a dose of 900 mg BID (BNC210.011; Table S8), steady state was achieved from Day 3, and Day 7 PK parameters were similar for males and females. Mean steady state AUC_0‐12 was 49.4 mg h/L at a dose of 900 mg BID (1800 mg/day), which is double the mean steady state AUC_0‐12 for 1000 mg BID administered as a suspension (24.9 mg h/L).

Pharmacodynamic Effects

Target Engagement

Since BNC210 is a negative allosteric modulator of the α7 nicotinic receptor, treatment with BNC210 was postulated to modify the response of the α2 band power to nicotine. In trial BNC210.005, a post‐hoc analysis was performed on a subgroup of 12 individuals in the BNC210 1,000 mg BID (2000 mg/day) treatment group who were identified as responders to nicotine prior to breaking the study blind. This analysis showed a significant difference in responses to doses of nicotine at 1 mg (P = 0.001), 1.5 mg (P = 0.001), and 2 mg (P = 0.020) between Day 1 (baseline) and Day 7, demonstrating a significant reduction in magnitude occurring after BNC210 dosing (Figure 2). Because BNC210 inhibited nicotine‐induced potential of the α2 band in this group of responders, these data suggest BNC210's target engagement of the α7 nAChR. The inhibition was partial, particularly with 2 mg nicotine, because BNC210, through its α7‐specific negative allosteric modulation, suppressed only the α7 activation, not the α4β2 activation, of the nAChRs in the CNS.

Figure 2.

BNC210 reduced nicotine‐induced quantitative wake EEG responses in the power of the α2 band after 7 days of 2000 mg/day dosing compared to pre‐dose, suggesting target engagement of the α7 nAChR (Trial BNC210.005). *P < 0.05, **P < 0.01, ***P < 0.001.

Neuroendocrine

No consistent effects of BNC210 were measured on serum cortisol or ACTH in these Phase 1 trials.

Subjective Feelings

No effects on subjective feelings were detected in four trials in healthy volunteers (BNC210.001, BNC210.002, BNC210.003, BNC210.005) when the participants were administered BNC210 in an unstimulated or basal state and asked to rate their feelings on a VAS (16‐item Bond and Lader or eVAS), whereas in one of these trials (BNC210.003), lorazepam 2 mg included in the trial as a positive control, showed a mood‐lowering effect on the eVAS at the 6 h time point (P < 0.05) compared to placebo (Figure 3, Panel A).

Figure 3.

Mean (±SEM) change from baseline effects following single dose administration of BNC210 (300 mg and 2000 mg) and lorazepam (2 mg, positive control) on pharmacodynamic parameters in healthy male volunteers (Trial BNC210.003). *P < 0.05, ***P < 0.001 versus placebo.

Cognitive Function

Clinical data supporting the absence of effects on attention and cognitive function with BNC210 come from the use of psychometric test batteries involving subjective and objective parameters in two key trials.

In the placebo‐ and lorazepam‐controlled trial (BNC210.003), 21 of the 24 participants received all four treatments in a cross‐over design and were included in the per‐protocol efficacy analysis set. BNC210 did not induce deleterious effects on attention and speed of processing, conversely to lorazepam:

• 1.In the Multiple Choice Reaction Time test, BNC210 showed no increases in total reaction time (TRT) at 300 or 2000 mg versus placebo at any time point, whereas TRT was significantly higher at 6 h (P < 0.001), 9 h (P < 0.001), and 12 h (P < 0.05) with lorazepam versus placebo (Figure 3, Panel B).
• 2.In the Digit Symbol Substitution Test, BNC210 showed no effects at any time point with the higher dose of 2000 mg and no effects with 300 mg, except at the 6‐h time point only (P < 0.05). Whereas lorazepam treatment showed a significant effect at the 6‐h (P < 0.001) and 9‐hour (P < 0.05) time points compared to placebo (Figure 3 Panel C).

Two of the tests (Critical Flicker Fusion Threshold and Perceptual Priming Test) were not sensitive to lorazepam in this trial, and therefore results from these two tests could not be used to draw valid conclusions on BNC210.

Using the PEPSY test battery in trial BNC210.005 on Day 1 and Day 8, BNC210 showed no overall consistent effect on cognition (no trend toward an improvement or impairment in performance). Although in a few instances an effect was observed in some of the tests, these effects were not dose‐related nor related to the day of dosing, and no consistency was found among the different cognitive functions tested at up to 1000 mg BID (2000 mg/day) for 8 days. There was no positive control in this trial. It is possible that the sporadic significant effects were due to the small sample size of the treatment groups, due to the models used for multiple statistical tests, and/or due to the large differences in baseline observed in some tests.

Sedation

The potential for BNC210 to be sedating was tested thoroughly in trial BNC210.003, which included lorazepam, a benzodiazepine with known sedative effects, as a positive control. No effects were observed for BNC210 on three different measurements of sedation:

• 1.On the Karolinska Sleepiness Scale, ¹⁵ compared to placebo, lorazepam caused a significant increase in the feeling of sleepiness at the 6‐h (P < 0.001) and 9‐h (P < 0.05) time points, whereas there was no increase in sleepiness with either BNC210 dose (Figure 3, Panel D).
• 2.Peak saccadic velocity (mean maximum) was significantly slower at the 6, 9, and 12 h (P < 0.001) with lorazepam compared to placebo, whereas peak saccadic velocity did not show any significant effect for either dose of BNC210 (Figure 3, Panel E).
• 3.Using qEEG, compared to placebo, lorazepam demonstrated increased delta activity during both the Resting and Vigilance‐Controlled conditions, which is typical for drugs with sedative effects, whereas BNC210 (300 and 2000 mg) increased delta EEG activity in the resting condition, but not the Vigilance‐Controlled condition. The lack of effect of BNC210 in the Vigilance‐Controlled condition supports a low likelihood of sedative effects and their reversal by a stimulus.

Abuse Liability

No clinically significant feelings associated with any classes of drugs of abuse were measured with BNC210 using the ARCI 49 checklist in three clinical trials (BNC210.003, BNC210.004, BNC210.005), suggesting that BNC210 has no or low potential for abuse. In one of these trials (BNC210.003), lorazepam 2 mg was included as a control, and on the ARCI 49 checklist, showed a significant effect on subjective sedation (P < 0.001) and dysphoric and psychotomimetic changes (P < 0.01) on the pentobarbital‐chlorpromazine‐alcohol and lysergic acid diethylamide subscales, respectively, and tended to score lower on the benzedrine subscale (P = 0.056), compared to placebo.

EEG Profiling

In trial BNC210.003, qEEG was used to determine drug‐induced changes in cerebral activity and showed a benzodiazepine profile for lorazepam with an increase in delta power, a decrease in theta and alpha power, and an increase in the whole beta‐band from 2.5 to 8.5 h post‐lorazepam dosing, in comparison to placebo. The modifications in EEG parameters were very robust since they were not only widespread in terms of good global effect over almost the whole scalp, but also of long duration, up to 8.5 h after administration. The qEEG changes fitted with hypovigilance (i.e., decreased attention associated with decreased alpha power) and sedation (i.e., increased delta power) as previously described in the literature. ²⁴ , ²⁵ , ²⁶ , ²⁷ BNC210 produced some increase in delta power in the Resting condition only (as described above), and a small decrease in alpha power (300 mg dose only), and only a slight increase in beta‐3. Overall, there was little overlap in EEG profiles between BNC210 and the positive control, lorazepam, and a low likelihood of sedation was postulated for BNC210.

Panic Symptoms

BNC210 demonstrated anti‐panic activity among the 15 evaluable participants who responded to the CCK‐4 challenge in trial BNC210.004. Pairwise analyses showed that at 10 min post CCK‐4 injection, participants treated with BNC210 2000 mg had a significant reduction in the Total Symptom score (total number of symptoms, P < 0.048) and in the Sum Intensity score (P < 0.041) on the PSS versus placebo‐treated participants (Figure 4). The eVAS was used to subjectively assess recovery to emotional stability following CCK‐4 injection, and among the 15 participants there was a trend (P = 0.104) toward improved emotional stability at 5 min post CCK‐4 injection with BNC210 compared to placebo (Figure 5).

Figure 4.

Mean (±SEM) change from baseline for (A) sum intensity score and (B) mean total symptom score on the panic symptom scale after a single 2000 mg dose of BNC210 versus placebo in 15 healthy male volunteers 10 min after a CCK‐4 challenge (Trial BNC210.004). *P < 0.05 versus placebo. CCK‐4, cholecystokinin tetrapeptide.

Figure 5.

Mean (±SEM) change from baseline emotional Visual Analogue Scale scores measuring recovery to emotional stability after a single 2000 mg dose of BNC210 in 15 healthy male volunteers following a CCK‐4 challenge (Trial BNC210.004). CCK‐4, cholecystokinin tetrapeptide.

Safety and Tolerability

No maximum tolerated dose was found for BNC210 in these eight Phase 1 trials that tested doses of up to 2000 mg with the suspension formulation and 1200 mg with the tablet formulation, and no serious adverse events were reported. Only one participant was withdrawn due to a treatment‐emergent adverse advent (TEAE; rib fracture, trial BNC210.005) and this was deemed not related to study treatment. The most frequently reported TEAEs are summarized in Table 2. While CNS effects such as headache, somnolence, nausea, and fatigue were reported in a greater proportion of the healthy volunteers on BNC210 compared to placebo, the SAD (BNC210.001, BNC210.002, BNC210.010) and MAD (BNC210.005) trials did not suggest a dose relationship for these events. Further, the psychometric tests performed in these Phase 1 trials and described above do not support a pattern of CNS‐related impairment. Beyond the Phase 1 trials reported here, there is a much larger safety database of patients who have been included in Phase 2 clinical trials, with treatment durations of up to 12 weeks. When BNC210 is administered to patients with GAD, social anxiety disorder, and PTSD, the frequency of reporting of CNS adverse events is balanced between BNC210 treatment and placebo, except for reports of somnolence, which are slightly higher for BNC210 (5.8%) versus placebo (2.0%) (Table S9).

Table 2.

Summary of the Most Frequently Reported Treatment Emergent Adverse Events in BNC210 Phase 1 Clinical Trials in Healthy Volunteers (Reported by >3% of Subjects)

	BNC210 (N = 173)	Placebo (N = 102)
Adverse event	n (%)	n (%)
Headache	35 (20.2)	6 (5.9)
Somnolence	16 (9.3)	0 (0)
Nausea	8 (4.6)	0 (0)
Fatigue	7 (4.1)	1 (0.9)

N, number of participants; n, number of participants reporting an AE.

Discussion

Dose selection and early proof‐of‐concept for a potentially first‐in‐class new treatment for psychiatric disease can be difficult. Key questions include whether the target has been engaged and whether downstream pharmacological effects have been produced. Adverse effects attributable to the CNS will be unwanted but may be a guide to the extent and time course of CNS penetration and action. However, being an anxiolytic compound, it was also important to demonstrate the lack of commonly experienced side effects of known anxiolytic compounds in the early stages of evaluation of BNC210. Further, the question of whether BNC210, as a negative allosteric modulator of α7 nAChR, would have any detrimental effects on cognition had to be answered. The results reported here from Phase 1 clinical trials back up the findings from the preclinical pharmacology and toxicology studies in various rodent models. ³

Poor absorption due to low aqueous solubility is a relatively common property of CNS‐directed orally administered therapies, because of the necessary lipophilicity to enter the CNS. A suspension with a small particle size giving high surface area is one potential solution to the problem, but absorption of BNC210 remained poor in the fasted state. A standardized high‐fat meal given under highly controlled conditions in a clinical research unit produced a large increase in absorption, and so it was recommended to patients in the first PTSD trial (NCT02933606) that the drug, administered as an aqueous suspension, be taken with food. However, it is likely that the effect of food on improving absorption for compounds with low aqueous solubility is related to the amount of fat in the meal. ²⁸ Patients would not typically be ingesting a meal of high fat content and may not reliably take the drug in a precise temporal relationship to food, and this may be a reason for the lower‐than‐expected plasma concentrations in the first PTSD trial. ⁹ That trial did not show the therapeutic benefit of BNC210, whereas a subsequent clinical trial (NCT04951076) with a similar design but using the spray‐dried dispersion tablet formulation, which achieved higher plasma concentrations, was effective. ⁸

A useful tool to assess target engagement in the brain is by PET; however no suitable ligand was available for BNC210. With BNC210 being a negative allosteric modulator, its pharmacological action may only be revealed under conditions of high stimulation to the receptor, which would not be the case in the basal state in resting, unstimulated, healthy volunteers. Two alternative methods were used to assess target engagement in healthy volunteers under stimulated conditions. One was by using a pharmacological stimulus of CCK‐4 induced panic, and the other was a more direct pharmacological model of examining the effects of nicotine on qEEG. Both of these studies showed effects of BNC210 consistent with its desired properties, giving confidence to proceed further in development.

Development of novel drugs in CNS indications is challenging, but here we describe a series of focused Phase 1 trials to explore the safety, tolerability, PK, and pharmacodynamic profile of BNC210. In these trials, BNC210 showed no CNS effects in unsimulated healthy volunteers, but whether in healthy volunteers or patients, pharmacological or fear induced activation of the relevant circuitry demonstrated the desired effects of BNC210, giving further support to proceed to definitive proof‐of‐concept clinical trials. Alongside the demonstration of pharmacodynamic effects in early clinical development of a drug, there is also the importance of developing a suitable formulation for later stage trials so that the drug's potential can be optimally evaluated.

Author Contributions

Conceptualization: Paul Rolan, Elizabeth Doolin, Dharam Paul, Philippe Danjou. Data curation: Paul Rolan, Elizabeth Doolin, Dharam Paul, Julia Crossman, Michael Odontiadis, Philippe Danjou. Formal analysis: Paul Rolan, Elizabeth Doolin, Dharam Paul, Julia Crossman, Michael Odontiadis, Philippe Danjou. Investigation: Paul Rolan, Elizabeth Doolin, Dharam Paul, Julia Crossman, Michael Odontiadis, Philippe Danjou. Methodology: Paul Rolan, Elizabeth Doolin, Dharam Paul, Julia Crossman, Michael Odontiadis, Philippe Danjou. Project administration: Paul Rolan, Elizabeth Doolin, Dharam Paul, Julia Crossman, Michael Odontiadis, Philippe Danjou. Supervision: Paul Rolan, Elizabeth Doolin, Dharam Paul, Michael Odontiadis, Philippe Danjou. Validation: Paul Rolan, Elizabeth Doolin, Dharam Paul, Julia Crossman, Michael Odontiadis, Philippe Danjou. Visualization: Paul Rolan, Elizabeth Doolin, Dharam Paul, Philippe Danjou. Writing: Paul Rolan, Elizabeth Doolin, Dharam Paul, Julia Crossman, Michael Odontiadis, Philippe Danjou, Mark Smith, Spyros Papapetropoulos.

Conflicts of Interest

All author are employees of or have received consulting fees from Neuphoria Therapeutics.

Funding

Funding was provided by Neuphoria Therapeutics Inc.

Supporting information

Supporting Information

Data Availability Statement

Further information about the data and conditions for access is available by contacting the corresponding author.