Effect of nivasorexant (ACT‐539313), a selective orexin‐1‐receptor antagonist, on multiple cytochrome P450 probe substrates in vitro and in vivo using a cocktail approach in healthy subjects

Abstract

Nivasorexant, a selective orexin‐1‐receptor antagonist, has recently been assessed in the treatment of humans with binge‐eating disorder. Herein, the inhibitory potential of nivasorexant on cytochromes P450 (CYPs) 2C9, 2C19, and 3A4 was evaluated. Human liver microsomes/recombinant CYP enzymes were evaluated in vitro. In vivo, a single‐center, open‐label, fixed‐sequence study was performed in healthy adults to explore the effect of 100 mg nivasorexant administered twice daily (b.i.d.) on the pharmacokinetics (PK) of flurbiprofen (50 mg, CYP2C9), omeprazole (20 mg, CYP2C19), midazolam (2 mg, CYP3A4) making use of a cocktail approach. Plasma PK sampling was performed over 24 h on Day 1 (Cocktail alone), 8 (Cocktail + nivasorexant), and 15 (Cocktail + nivasorexant at steady state). Genotyping of subjects' CYPs was performed while safety and tolerability were also assessed. In vitro, nivasorexant inhibited CYP2C9, 2C19, and 3A4 in competitive inhibition assays with IC₅₀ values of 8.6, 1.6, and 19–44 μM, respectively, while showing a significant time‐dependent CYP2C19 inhibition. In 22 subjects, exposure to flurbiprofen, omeprazole, and midazolam was generally higher during concomitant single‐ (i.e., area under the plasma concentration–time curve [AUC] ratio increased by 1.04‐, 2.05‐, and 1.56‐fold, respectively) and repeated‐dose (i.e., AUC ratio increased by 1.47‐, 6.84‐, and 3.71‐fold, respectively) nivasorexant administration compared with the cocktail substrates administered alone. The most frequently reported adverse event was somnolence. According to regulatory guidance, nivasorexant is classified as a moderate CYP2C19 and weak CYP3A4 inhibitor after 1 day and as a weak CYP2C9, strong CYP2C19, and moderate CYP3A4 inhibitor after 8 days of 100 mg b.i.d. administration. Clinicaltrials.gov ID: NCT05254548.

Arithmetic mean (+SD) plasma concentration‐time profiles (linear and semi‐logarithmic scales) of cocktail substrates (i.e., flurbiprofen, omeprazole, midazolam, and their CYP isoenzyme‐specific metabolites) after administration of the cocktail alone and with nivasorexant.

Abbreviations

AE

adverse event

AIX

accumulation index

AUC

area under the plasma concentration‐time curve

AUC_0–24

area under the plasma concentration‐time curve from zero to time 24 h

AUC_τ

area under the plasma concentration‐time curve during a dose interval

b.i.d.

twice daily

BED

binge‐eating disease

BMI

body mass index

CI

confidence interval

C _max

maximum plasma concentration

CYP

cytochrome P450

DDI

drug–drug interaction

GM(R)

geometric mean (ratio)

IC₅₀

concentration that causes 50% inhibition

ISP

ion sphere particle

K _i

inhibition constant

K _I

inhibition constant of the CYP inactivation

K _inact

rate of enzyme inactivation

K _m

Michaelis–Menten constant

LLOQ

lower limit of quantification

OH

hydroxy

OX1R

orexin receptor type 1 receptor

OX2R

orexin receptor type 2 receptor

PCR

polymerase chain reaction

PK

pharmacokinetic(s)

POC

proof‐of‐concept

SO1RA

selective orexin‐1‐receptor antagonist

t _½

terminal half‐life

t _max

time to reach maximum plasma concentration

1. INTRODUCTION

Nivasorexant (formerly referred to as ACT‐539313) is a potent, selective, and brain‐penetrating orexin receptor type 1 receptor (OX1R) antagonist with an approximately 60‐fold higher potency at human OX1R (apparent equilibrium dissociation constant [Kb]: 0.69 nM) than at human orexin receptor type 2 receptor (OX2R; apparent Kb: 42 nM). ¹ OX1R appears to be primarily important for orexins' effects on the reward and stress systems, whereas OX2R is mostly responsible for orexins' effects on wakefulness. ² , ³ The effects of OX1R antagonism on various food‐ and drug‐reward‐related behaviors, as well as on anxiety and stress responses, in animals supported the potential benefit of OX1R antagonists in patients with eating disorders, especially binge‐eating disease (BED). ⁴ , ⁵ , ⁶ , ⁷

The clinical development program for this selective OX1R antagonist (SO1RA) nivasorexant includes two Phase 1 studies in healthy subjects, i.e., a single‐center, double‐blind, randomized, placebo‐controlled, single−/multiple‐ascending and exploratory pharmacodynamic Phase 1 study as well as a single‐center, open‐label, drug–drug interaction (DDI) study on cytochrome P450 (CYP) 3A4 inhibition. ¹ , ⁸ , ⁹ Furthermore, a Phase 2 proof‐of‐concept (POC) study in patients with moderate to severe BED was completed at the same time as the present study. The POC study assessed nivasorexant 100 mg b.i.d. versus placebo in a 12‐week study being the first to explore this new mechanism of action in a psychiatric condition with the aim of establishing an initial signal of efficacy and to ensure the drug's safety (Clinicaltrials.gov ID: NCT04753164).

The PK of nivasorexant were thoroughly investigated in the aforementioned first‐in‐human study, wherein the drug was administered as a single dose, that is, in the dose range of 10–400 mg, as well as at doses of 30, 100, and 200 mg b.i.d. for up to 6.5 consecutive days. ¹ , ⁹ Nivasorexant was shown to be rapidly absorbed with a median time to maximum plasma concentration (t _max) of 0.7 to 3.5 h and with an apparent mean terminal half‐life (t _½) of 3.3 to 6.5 h across dose levels in healthy male and female subjects. Following multiple oral administration, the area under the plasma concentration–time curve (AUC) during a dosing interval (AUC_τ) was dose proportional, while maximum plasma concentration (C _max) was less than dose proportional. Furthermore, in the previously conducted DDI study, it was shown that the PK of midazolam, a sensitive substrate of CYP3A4 through which its main metabolite, that is, 1‐hydroxy (OH) midazolam, is formed, were affected whereby an increase in AUC of 1.8‐and 4.5‐fold was observed following initial and repeated b.i.d. dosing (over 10 days) with 200 mg nivasorexant. ⁹ , ¹⁰

The objective of this body of work is to present the inhibition potential of nivasorexant on CYPs, as assessed in vitro and in a further clinical study. Prior to the conduct of the clinical study, the most pronounced inhibition of nivasorexant in vitro was observed on CYP2C19, and to a lesser extent, on CYP2C9, while assays have shown the microsomal metabolism of nivasorexant in man to be dependent on CYP3A4 to an extent of approximately 90%. ⁹ Thus, as the effect of nivasorexant on both CYP2C9 and CYP2C19 in a clinical setting was unknown, the clinical study aimed to simultaneously investigate its effect on the PK of flurbiprofen and omeprazole, sensitive substrates of CYP2C9 and CYP2C19, respectively, as well as its effect on the PK of midazolam at a lower dose of nivasorexant than administered previously (i.e., at 100 mg b.i.d.). ⁹ Flurbiprofen is metabolized by CYP2C9 to 4‐OH flurbiprofen, while omeprazole is metabolized by CYP2C19 to 5‐OH omeprazole. ¹¹ , ¹² , ¹³ They are both used experimentally as probe drugs of CYP2C9 and CYP2C19, respectively. The simultaneous investigation, known as a so called “cocktail approach,” of the DDI potential of a perpetrator compound on the PK of the respective cocktail substrates of interest is in line with regulatory guidelines and is possible due to the documented lack of mutual interactions among the three probe drugs and as it is feasible to fully characterize the plasma concentration–time curves of flurbiprofen, omeprazole, midazolam, and their CYP isoenzyme‐specific hydroxymetabolites within a common time course. ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸

2. METHODS

2.1. Chemicals and reagents used in the in vitro CYP inhibition assays

Nivasorexant with a molecular weight of 429.5 g/mol and the deuterated analogue ACT‐539313C were produced by Actelion Pharmaceuticals Ltd, the predecessor of Idorsia Pharmaceuticals Ltd. All other chemicals and solvents used throughout this study were of highest commercially available quality and obtained from Lipomed, Sigma‐Aldrich, Santa Cruz Biotechnology, ALSA Chim, and Corning Life Sciences. Pooled human liver microsomes as well as recombinant CYP2C19 expressed in baculovirus‐infected insect cells with supplemental cDNA‐expressed P450 reductase and cytochrome b5 were purchased from Corning Life Sciences.

2.2. In vitro CYP inhibition assays

The effect of nivasorexant on the competitive (reversible) inhibition of CYPs was tested using human liver microsomes (except for CYP2C19 for which recombinant CYP enzyme expressed in baculovirus‐infected Sf9 cells were used) and CYP isoform‐specific marker transformations as shown in Table 1. Two different probes, that is, midazolam and testosterone were used to assess the inhibitory potential of nivasorexant on CYP3A4 activity. To determine the concentration that causes 50% inhibition (IC₅₀), marker substrates were used at concentrations around their respective Michaelis–Menten constant (K _m) values, while nivasorexant concentrations of up to 100 μM were used. The inhibitory effect was assessed by quantifying metabolite formation by liquid chromatography–tandem mass spectrometry (LC/MS‐MS) using qualified methods, calibration curves, and quality control samples. The potential of nivasorexant to cause changes in the enzymatic activity of CYP2C8, CYP2C9, CYP2C19, CYP2D6, and CYP3A4 due to time‐dependent (irreversible) inhibition was also investigated using either human liver microsomes or recombinant CYP2C19. The marker substrates were used at a single concentration approximately 10‐fold higher than their respective K _m values. Two independent sets of experiments were performed in which the effect of nivasorexant concentrations up to 100 μM was tested on each individual CYP. In the first experiment with a pre‐incubation period in the absence of nicotinamide adenine dinucleotide phosphate (NADPH), NADPH was added to start the reaction when both nivasorexant and the marker substrate were present. In the second experiment with a pre‐incubation in the presence of NADPH, NADPH was added at the same time as nivasorexant. The marker substrate was added only at the end of the 30 min pre‐incubation period. IC₅₀ values were determined from both experiments, and the ratio of the two IC₅₀ values was calculated. Thus, the so‐called IC₅₀ shift, a marker of the extent of change in the enzyme activity during the pre‐incubation period, was calculated with which it is possible to distinguish between reversible and irreversible inhibition. Curve fitting for IC₅₀ determinations was performed and plots for inhibition constant (K _i) determinations were generated, where appropriate, using GraphPad Prism (version 7.02, GraphPad Software Inc.). K _i values were determined using non‐linear regression analysis and the mixed model in the GraphPad Prism software, while curve fitting for the determination of the rate of enzyme inactivation (k _inact) and the inhibition constant of the CYP inactivation (K _I) was performed using the same software.

TABLE 1.

Inhibition of human CYP enzymes by nivasorexant in vitro.

CYP isoform	Marker transformation	Competitive inhibition assay		Time‐dependent inhibition assay
		IC50 (μM)	K i (μM)	IC50 shift	K I (μM)	k inact (min−1)
CYP1A2	Phenacetin‐O‐deethylation	> 100a	n.d.	none	n.d.	n.d.
CYP2A6	Coumarin 7‐hydroxylation	> 100b	n.d.	none	n.d.	n.d.
CYP2B6	Bupropion hydroxylation	> 100c	n.d.	none	n.d.	n.d.
CYP2C8	Paclitaxel 6α‐hydroxylation	25	n.d.	none	n.d.	n.d.
CYP2C9	Diclofenac 4′‐hydroxylation	8.6	5.0	none	n.d.	n.d.
CYP2C19	(S)‐mephenytoin 4′‐hydroxylation	1.6	0.8	67	4.2	0.079
CYP2D6	Dextromethorphan N‐demethylation	> 100d	n.d.	none	n.d.	n.d.
CYP3A4	Midazolam 1′‐hydroxylation	19	n.d.	none	n.d.	n.d.
CYP3A4	Testosterone 6β‐hydroxylation	44	n.d.	none	n.d.	n.d.

Note: ^a23%, ^b19%, ^c44%, ^d33% inhibition at the highest concentration of 100 μM nivasorexant, respectively.

Abbreviations: CYP, cytochrome P450; IC₅₀, concentration that causes 50% inhibition; K _i, inhibition constant; K _I, inhibition constant of the CYP enzyme inactivation; k _inact, rate of enzyme inactivation; n.d., not determined.

2.3. Clinical study

The study was approved by the German Health Authorities (BfArM) and by the local ethics committee (Landesamt für Gesundheit und Soziales). The study adhered to the Declaration of Helsinki and was conducted at CRS Berlin GmbH according to good clinical practice guidelines, and applicable laws and regulations. Prior to any study procedure, written informed consent was obtained from each participant after adequate explanation of the objectives, methods, and potential hazards of the study.

2.4. Subjects and treatment

Non‐smoking, healthy female, and male subjects, aged between 18 and 45 years with a body mass index (BMI) between 18.5 and 28.0 kg/m² (inclusive), were recruited in this study. Eligibility of the study participants was assessed at a screening examination, which took place between 3 and 28 days prior to first drug administration. The subjects were in good health, as assessed by medical history, physical examination, 12‐lead electrocardiogram (ECG), and clinical laboratory tests (including hematology, clinical chemistry, urinalysis, virus serology, and drug screening). Vital signs had to be within the normal range, defined as 100–140 mmHg for systolic pressure, 50–90 mmHg for diastolic pressure, and 45–90 bpm for pulse rate. Women of childbearing potential had to have a negative serum pregnancy test at screening, a negative urine pregnancy test on Day −1, and had to consistently and correctly use (from screening, during the entire study, and for at least 30 days after last study treatment intake) a highly effective method of contraception with a failure rate of < 1% per year, be sexually inactive, or have a vasectomized partner. Consumption of food and beverages that could influence CYP activity (e.g., grapefruit) as well as xanthine‐containing beverages, nicotine, and alcohol was forbidden. Concomitant therapy (except for ibuprofen [1200 mg/day]/paracetamol [up to 1500 mg/day] up until Day −1 or contraceptives) was prohibited unless required for the treatment of an adverse event (AE).

This was a single‐center, open‐label Phase 1 study (Clinicaltrials.gov ID: NCT05254548), enrolling 22 healthy female and male subjects (in a 1:1 ratio) to assess the effect of single‐ and multiple‐dose nivasorexant on the PK of flurbiprofen, omeprazole, midazolam, and their respective hydroxymetabolites. For the investigation of CYP2C9, CYP2C19, and CYP3A4 inhibition by nivasorexant, the probe cocktail (i.e., 50 mg flurbiprofen [tablet], 20 mg omeprazole [gastro‐resistant capsule], 2 mg midazolam [oral solution]) was given alone on Day 1, 1 h after the first administration of 100 mg nivasorexant on Day 8 as well as following 7 days of repeated 100 mg b.i.d. dosing to evaluate the CYP inhibition by nivasorexant at steady state (on Day 15), see Figure 1. ⁸ , ⁹ To be able to consecutively investigate the effect of nivasorexant on the CYPs of interest upon first dose of 100 mg b.i.d. as well as following repeated dosing with 100 mg b.i.d., a sequential design was chosen. The safety and tolerability of nivasorexant administered alone and in combination with the probe cocktail were also assessed.

FIGURE 1.

Study design. Administration of the cocktail and nivasorexant under fasted conditions in the morning and outside of meal times in the evening only on Days 1, 8, and 15. AV ambulatory visit, b.i.d. twice daily, cocktail 50 mg flurbiprofen +20 mg omeprazole +2 mg midazolam, EOP end‐of‐period, EOS end‐of‐study, SFU safety follow‐up (30–40 days after EOS), SCR screening (from Day −28 to Day −10 for women of childbearing potential and between Day −28 and − 3 for all other subjects), Treatment A cocktail administered alone in the morning on Day 1, Treatment B concomitant administration of the cocktail with initial dosing of 100 mg nivasorexant (i.e., cocktail administered 1 h after initial administration of nivasorexant) in the morning on Day 8 (followed by evening administration of nivasorexant on Day 8), Treatment C 100 mg b.i.d. nivasorexant administered from the morning of Day 9 until the evening of Day 14, Treatment D concomitant administration of the cocktail with 100 mg nivasorexant in the morning (followed by last 100 mg nivasorexant dose administered in the evening of Day 15).

2.5. Blood sampling and bioanalysis

Blood samples were collected from a vein in the cubital or antebrachial region through an indwelling catheter or by direct venepuncture. Serial blood samples were collected in tubes containing K ₃‐EDTA on Days 1, 8, and 15 pre‐dose and regularly until 24 h after probe cocktail intake to assess the PK profiles of the probe substrates and their respective hydroxymetabolites with trough (pre‐dose) plasma concentrations of nivasorexant collected in the mornings from Day 9–14. After centrifugation, plasma was transferred to polypropylene tubes and stored in an upright position at – 20°C or below.

Plasma nivasorexant concentrations were determined using a previously described validated LC‐MS/MS assay with a lower limit of quantification (LLOQ) of 1.0 ng/mL and covering a range up to 3000 ng/mL. ¹ The inter‐assay accuracy was between −1.3% and 2.1% and the inter‐assay precision was ≤5.4%. Two further validated LC‐MS/MS assays were used to measure plasma concentrations of flurbiprofen and 4‐OH flurbiprofen (method 2) as well as omeprazole, 5‐OH omeprazole, midazolam, and 1‐OH midazolam (method 3). The LLOQ was 10 ng/mL for flurbiprofen and 4‐OH flurbiprofen, 0.5 ng/mL for omeprazole and 5‐OH omeprazole, and 0.1 ng/mL for midazolam and 1‐OH midazolam. For method 2, following liquid–liquid extraction, chromatographic separation was achieved using a reverse phase pre‐column (Security Guard Phenyl, 4 × 3 mm, Phenomenex) and a Luna Phenyl‐Hexyl column (100 Å, 5 μm, 150 × 4.6 mm, Phenomenex) with an injection volume of 10.0 μL (PAL HTC‐xt autosampler, CTC Analytics AG) at a flow rate of 0.800 mL/min and mobile phases consisting of water containing 1 mM ammonium formate and acetonitrile (LC‐20 AD pumps, Shimadzu). Mass spectrometric detection was performed with a triple quadrupole mass spectrometer (Triple Quad 5500, AB Sciex, Concord) operating in electrospray ionization mode. Flurbiprofen, flurbiprofen‐d3, 4‐OH flurbiprofen, 4‐OH flurbiprofen‐d3 transitions, i.e., parent m/z > fragment m/z, were 242.9 > 199.0 amu, 247.0 > 202.8 amu, 258.4 > 214.9 amu, and 263.0 > 219.0 amu, respectively. The statistics of the QC samples for flurbiprofen and 4‐OH flurbiprofen showed that the inter‐run accuracy was in the range from −0.8% to 7.1% whereas the inter‐assay precision was ≤4.8%. For method 3, chromatographic separation was achieved by column separation with reverse phase chromatography. Thereby, on‐line solid phase extraction with a reverse phase ReproSil Gold C18 trapping column (10 × 2 mm, 5 μm, Dr. Maisch HPLC GmbH) and a YMC Pro C4 analytical column (2.1 × 50 mm, 3 μm, YMC Co. Ltd.) with an injection volume of 10.0 μL (CTC PAL autosampler, CTC Analytics AG) at a flow rate of 0.600 mL/min and mobile phases consisting of water containing 5 mM ammonium formate and acetonitrile (1200 series high‐performance liquid chromatography pump, Agilent Technologies Inc) was used. A triple‐stage quadrupole Vantage mass spectrometer (Thermo Fisher Scientific) was used operating in selective reaction monitoring mode. Omeprazole, omeprazole‐d3, 5‐OH omeprazole, 5‐OH omeprazole‐d3, midazolam, midazolam‐d4, 1‐OH midazolam, 1‐OH midazolam‐d4 transitions were 346.1 > 198.1 amu, 349.4 > 198.1 amu, 362.1 > 214.1 amu, 365.1 > 214.1 amu, 326.1 > 291.1 amu, 330.1 > 295.1 amu, 342.1 > 203.1 amu, and 346.1 > 203.1 amu, respectively. The statistics of the QC samples for the analytes measured in method 2 showed that the inter‐run accuracy was in the range from −3.5% to 2.6% whereas the inter‐assay precision was ≤11.7%.

2.6. Pharmacokinetic evaluation and statistics

The PK parameters of flurbiprofen, omeprazole, midazolam and their respective hydroxymetabolites were obtained by non‐compartmental analysis using Phoenix WinNonlin (version 8.3.3.33; Certara). The measured individual plasma concentrations of each analyte were used to directly obtain C _max and t _max. AUC from time 0 to 24 h after dosing (AUC_0–24) and AUC_τ (for nivasorexant only) were calculated according to the linear trapezoidal rule. The accumulation index (AIX) was calculated by comparing the AUC_τ between treatment days (i.e., AUC_τ on Day 15/ AUC_τ on Day 8). The terminal half‐life values were calculated as follows: t _½ = ln (2)/λ _z, whereby λ _z represents the terminal elimination rate constant determined by log‐linear regression analysis of the measured plasma concentrations of the terminal elimination phase. Assessment of steady‐state concentrations of nivasorexant was based on pre‐dose trough concentration data and confirmed statistically by Helmert contrasts.

The sample size of 22 subjects was based on the application of a precision estimate approach for AUC_0–24 and C _max comparisons of omeprazole, that is, the probe drug with the highest variability in PK. ¹⁹ PK variables were analyzed providing geometric mean (GM) and corresponding 95% confidence intervals (CI) for AUC_0–24, C _max, and t _½ whereas the median and range were used for t _max. The effect of nivasorexant on AUC_0–24, C _max, and t _½ of the respective probe substrates and their metabolites was explored using the GM ratio (GMR) and 90% CIs of Treatment B (Day 8, cocktail + initial‐dose nivasorexant) or Treatment D (Day 15, cocktail + repeated‐dose nivasorexant b.i.d., i.e., at steady state) with Treatment A (Day 1, cocktail alone) as reference treatment. The log‐transformed values were analyzed by mixed‐effect models including treatment, sequence, and period as fixed effects and subject as a random effect. Differences between treatments for t _max were explored using the Wilcoxon signed‐rank test providing the median differences and corresponding 90% CI. All statistical analyses were performed using SAS® version 9.4 (SAS Institute).

2.7. Genotype analysis

A blood sample of 2.7 mL was taken from each subject (predose on Day 1) with which the genetic polymorphisms of CYP2C9, CYP2C19, and CYP3A4 were evaluated. Therefore, DNA was isolated from EDTA blood on a QIAsymphony instrument (Qiagen). Primer panel and multiplex‐polymerase chain reaction (PCR) amplicon libraries were generated from genomic DNA. The amplicons covered all exons of the gene set, and in addition non‐exonic single‐nucleotide polymorphisms. The amplicon libraries were ligated with adapters, bar coded on the molecular level, and bound to ion sphere particles (ISPs). ISP‐bound templates were then monoclonaly amplified using emulsion‐PCR and sequenced on an Ion GeneStudio S5 System (Thermo Fisher Scientific) with semiconductor‐sequencing technology (next‐generation sequencing). The generated raw sequence data were analyzed with regard to coverage depth and frequency of allelic variants of the tested genes (i.e., alleles *1, *2, *3, *4, *5, *6, *8, *11, *14, *17, *24, and *26 for CYP2C9 and *1, *2, *3, *4, *5, *6, *7, *8, *9, *10, and *17 for CYP2C19). CYP3A4 (allele *22) was genotyped by conventional Sanger sequencing using an ABI3130 genetic analyzer (Applied Biosystems).

2.8. Safety assessments

Safety and tolerability were assessed based on vital sign measurements, 12‐lead ECG, physical examination, hematology, clinical chemistry, urinalysis, and AE data.

3. RESULTS

3.1. CYP inhibition in vitro

Nivasorexant exhibited weak inhibition of CYP1A2, CYP2A6, CYP2B6, and CYP2D6, with IC₅₀ values in excess of 100 μM (see Table 1). More pronounced inhibition was observed on all members of the CYP2C family, with nivasorexant inhibiting CYP2C8, CYP2C9, and CYP2C19 in the competitive inhibition assays with IC₅₀ values of 25, 8.6, and 1.6 μM, respectively (Table 1). The corresponding K _i values were 5.0 and 0.8 μM for CYP2C9 and CYP2C19, respectively. Weaker inhibition was observed for CYP3A4 with IC₅₀ values of 19 and 44 μM, using midazolam and testosterone as different probe substrates. In the time‐dependent inhibition assays, a significant IC₅₀ shift of 67 was observed on CYP2C19, while pre‐incubation with nivasorexant had no effect on CYP2C8, CYP2C9, CYP2D6, or CYP3A4 activity. For CYP2C19, enzyme inactivation parameters, K _I and k _inact were determined as 4.2 μM and 0.079 min⁻¹, respectively.

3.2. Clinical study

3.2.1. Demographics

Twenty‐two healthy subjects (11 female and 11 male), with a mean age of 36.6 years (range, 24–45 years) and a mean BMI of 24.2 kg/m² (range, 20.9–28.0 kg/m²) were enrolled in this study. Twenty‐one subjects were white, while one subject was of Asian descent. Results of the genotype analysis showed that one subject was a poor metabolizer of CYP2C9 (i.e., with a combination of a decreased function and no function allele [CYP2C9 *2/*3]) while another subject was a poor metabolizer of CYP3A4 (i.e., homozygous for a reduced function allele, CYP3A4*22/*22). Both identified poor metabolizers were kept in the PK set but not included in the analysis of the effect of nivasorexant on the respective probe substrates. Furthermore, eight subjects for CYP2C9 and five subjects for CYP2C19 were classified as intermediate metabolizers, respectively, while for CYP2C19, seven subjects were identified as rapid metabolizers. All remaining subjects were identified as normal metabolizers carrying wild‐type alleles for CYP2C9 (i.e., CYP2C9 *1/*1 [N = 13]), CYP2C19 (i.e., CYP2C19 *1/*1 [N = 10]), and CYP3A4 (i.e., CYP3A4 *1/*1 [N = 21]), respectively. In Supplemental Figure S1, the flurbiprofen and midazolam concentration‐time profiles of the excluded subjects with CYP2C9 *2/*3 (N = 1) and CYP3A4*22/*22 (N = 1) are shown in comparison to the subjects included in the analysis of CYP2C9 and CYP3A4 by treatment (i.e., either intermediate metabolizers [for CYP2C9 only] or normal metabolizers carrying wild‐type alleles [N = 21]), while the differing PK parameters are shown in Supplemental Table S1.

3.2.2. Pharmacokinetics

Nivasorexant

After initial (on Day 8, Treatment B) and following repeated (on Day 15, Treatment D) administration of 100 mg b.i.d. nivasorexant, plasma concentration–time profiles of nivasorexant over one dosing interval were comparable (see Figure 2). nivasorexant was quickly absorbed with a median t _max of 3.0 h and 2.0 h, respectively (see Table 2). Thereby, C _max values of 693 ng/mL and 1252 ng/mL were reached, while AUC_τ was 4686 ng·h/mL and 8558 ng·h/mL on Day 8 and Day 15, respectively. Mean trough nivasorexant concentrations reached steady‐state conditions after 3–4 days of b.i.d. treatment. After 8 days of treatment with 100 mg b.i.d., nivasorexant accumulated with an AIX of 1.8.

FIGURE 2.

Arithmetic mean (+SD) plasma concentration‐time profiles (linear and semi‐logarithmic scales) over the first dosing interval of nivasorexant following initial administration of 100 mg on Day 8 (Treatment B) and following repeated 100 mg b.i.d. nivasorexant administration on Day 15, i.e., at steady state (Treatment D). The calculated pharmacokinetic parameters are given in Table 2. b.i.d., twice daily; SD, standard deviation.

TABLE 2.

Pharmacokinetic parameters of flurbiprofen, omeprazole, midazolam, their respective CYP isoenzyme‐specific hydroxymetabolites, and nivasorexant.

Parameter [unit]	Trt A [N = 22 or 21] a	Trt B [N = 22 or 21] a	Trt D [N = 21 or 20] a	Trt B/ Trt A b	Trt D/ Trt A b
Flurbiprofen
C max [ng/mL]	6414 (5704–7212)	6194 (5730–6695)	6954 (6250–7737)	0.97 (0.86–1.08)	1.08 (0.97–1.20)
t max [h]	2.00 (0.75–6.00)	3.00 (1.00–6.02)	2.50 (0.75–6.00)	0.75 (0.00–1.51)	0.50 (−0.13–1.16)
AUC0–24 [ng*h/mL]	36 160 (33662–38 844)	37 513 (34745–40 501)	52 996 (47479–59 155)	1.04 (1.00–1.08)	1.47 (1.39–1.55)
t ½ [h]	5.73 (5.40–6.08)	6.36 (5.97–6.77)	7.89 (7.25–8.58)	1.11 (1.07–1.15)	1.39 (1.31–1.47)
4‐OH flurbiprofen
C max [ng/mL]	316 (278–359)	289 (259–322)	175 (157–195)	0.91 (0.87–0.96)	0.53 (0.49–0.59)
t max [h]	3.00 (1.50–6.00)	3.00 (1.50–6.02)	3.00 (1.50–6.00)	0.74 (−0.02–1.50)	0.50 (−0.04–1.00)
AUC0–24 [ng*h/mL]	2516 (2280–2778)	2366 (2138–2618)	1924 (1763–2099)	0.94 (0.90–0.98)	0.74 (0.70–0.79)
t ½ [h]	7.85 (7.04–8.75)	8.03 (7.28–8.87)	12.0 (10.0–14.3)	1.02 (0.97–1.09)	1.54 (1.39–1.70)
Omeprazole
C max [ng/mL]	162 (118–223)	260 (203–334)	566 (473–679)	1.60 (1.30–1.97)	3.31 (2.55–4.28)
t max [h]	3.00 (1.00–6.00)	3.49 (1.50–6.02)	3.00 (1.50–10.0)	0.50 (0.00–1.49)	1.00 (0.50–1.75)
AUC0‐24 [ng*h/mL]	372 (266–522)	763 (549–1059)	2667 (2283–3114)	2.05 (1.80–2.33)	6.84 (5.35–8.74)
t ½ [h]	0.99 (0.81–1.21)	1.45 (1.14–1.84)	1.99 (1.74–2.28)	1.46 (1.24–1.72)	2.01 (1.69–2.40)
5‐OH omeprazole
C max [ng/mL]	129 (110–151)	97.0 (85.0–110.7)	42.2 (37.9–47.0)	0.75 (0.66–0.86)	0.53 (0.49–0.59)
t max [h]	3.00 (1.00–6.00)	3.49 (1.50–6.02)	4.00 (1.50–10.00)	0.50 (0.00–1.16)	1.25 (0.75–2.00)
AUC0–24 [ng*h/mL]	367 (332–406)	338 (314–363)	278 (253–305)	0.92 (0.85–1.00)	0.74 (0.67–0.82)
t ½ [h]	1.34 (1.20–1.49)	1.53 (1.30–1.80)	2.03 (1.79–2.32)	1.14 (0.99–1.32)	1.50 (1.36–1.66)
Midazolam
C max [ng/mL]	6.13 (5.23–7.19)	8.20 (7.05–9.54)	16.4 (14.4–18.7)	1.34 (1.22–1.46)	2.62 (2.31–2.97)
tmax [h]	0.75 (0.50–3.00)	0.50 (0.50–1.50)	0.50 (0.50–1.50)	0.00 (−0.13–0.13)	0.00 (−0.13–0.13)
AUC0–24 [ng*h/mL]	14.6 (12.7–16.7)	22.7 (19.5–26.5)	54.8 (45.5–66.1)	1.56 (1.43–1.70)	3.71 (3.20–4.29)
t ½ [h]	4.94 (4.02–6.06)	8.24 (6.78–10.02)	6.75 (6.00–7.60)	1.67 (1.38–2.01)	1.25 (1.07–1.46)
1‐OH Midazolam
C max [ng/mL]	4.03 (3.23–5.02)	4.12 (3.53–4.82)	3.55 (2.79–4.51)	1.02 (0.91–1.15)	0.86 (0.75–0.98)
t max [h]	0.50 (0.50–3.00)	0.75 (0.50–1.50)	0.50 (0.50–1.50)	0.00 (0.000–0.13)	0.00 (−0.13–0.13)
AUC0–24 [ng*h/mL]	8.50 (7.23–10.00)	8.64 (7.37–10.13)	9.88 (8.02–12.18)	1.02 (0.96–1.08)	1.15 (1.06–1.24)
t ½ [h]	2.27 (1.58–3.25)	2.94 (2.04–4.22)	9.04 (6.82–11.97)	1.29 (1.03–1.62)	3.05 (2.03–4.58)
Nivasorexant
C max [ng/mL]		693 (619–777)	1252 (1094–1433)
t max [h]		3.00 (0.95–7.00)	2.00 (0.50–6.00)
AUCτ [ng*h/mL]		4686 (4226–5197)	8558 (7143–10 252)

Abbreviations: AUC_0–24, area under the plasma concentration–time curve from time zero to 24 h; AUC_τ area under the plasma concentration–time curve during a dosing interval; CI, confidence interval; C _max maximum plasma concentration; h, hours; N, number of subjects; t _½, terminal half‐life; t _max, time to reach maximum plasma concentration.

^a Data are geometric means (and 95% CI) or for t _max the median (and range).

^b Data are ratios of geometric means (B/A, D/A; and 90% CI) or for t _max the median differences (B–A, D–A; and 90% CI).

Flurbiprofen and 4‐OH flurbiprofen

Exposure to flurbiprofen and 4‐OH flurbiprofen was unchanged following concomitant initial administration of 100 mg b.i.d. nivasorexant on Day 8 (Treatment B) compared with flurbiprofen administered alone (Treatment A, see Figure 3A,B and Table 2). GMRs of C _max, AUC_0–24, and t _½ between Treatment B and Treatment A of both analytes were comparable with values ranging from 0.91 to 1.11. However, on Day 15, that is, on the 8th day of nivasorexant b.i.d. dosing (Treatment D), the greater exposure to flurbiprofen was inversely mirrored by a decreased exposure to 4‐hydroxyflurbiprofen (see Figure 3A,B and Table 2). Thereby, GMRs (90% CI) between Treatment D and Treatment A of C _max, AUC_0–24, and t _½ were 1.08 (0.97–1.20), 1.47 (1.39–1.55), and 1.39 (1.31–1.47), respectively, for flurbiprofen, and 0.53 (0.49–0.59), 0.74 (0.70–0.79), and 1.54 (1.39–1.70), respectively, for 4‐OH flurbiprofen. Difference in median t _max was only minimal.

FIGURE 3.

Arithmetic mean (+SD) plasma concentration‐time profiles (linear and semi‐logarithmic scales) of flurbiprofen (A), omeprazole (C), and midazolam (E) as well as their CYP isoenzyme‐specific metabolites 4‐OH flurbiprofen (B), 5‐OH omeprazole (D), and 1‐OH midazolam (F) after administration of the cocktail alone (Treatment A, Day 1), concomitantly with an initial dosing of 100 mg b.i.d. nivasorexant (Treatment B, Day 8), and concomitantly following repeated 100 mg b.i.d. nivasorexant administration, that is, at steady state (Treatment D, Day 15). The calculated pharmacokinetic parameters are given in Table 2. b.i.d., twice daily, SD, standard deviation.

Omeprazole and 5‐OH omeprazole

Following concomitant administration with the initial 100 mg b.i.d. nivasorexant dose on Day 8, omeprazole and 5‐OH omeprazole PK parameters showed a distinct change compared with administration of omeprazole alone in Treatment A (see Figure 3C,D and Table 2). GMRs (90% CI) between Treatment B and Treatment A of C _max, AUC_0–24, and t _½ were 1.60 (1.30–1.97), 2.05 (1.80–2.33), and 1.46 (1.24–1.72), respectively, for omeprazole, and 0.75 (0.66–0.86), 0.92 (0.85–1.00), and 1.14 (0.99–1.32), respectively, for 5‐OH omeprazole. T _max was slightly delayed with a median difference (90% CI) of 0.5 h (0.00–1.49). Exposure to omeprazole was even higher in Treatment D, whereby GMRs (90% CI) compared with Treatment A of C _max, AUC_0–24, and t _½ were 3.31 (2.55–4.28), 6.84 (5.35–8.74), and 2.01 (1.69–2.40), respectively, for the parent and 0.53 (0.49–0.59), 0.74 (0.67–0.82), and 1.50 (1.36–1.66), respectively, for its 5‐OH metabolite (see Figure 3C,D and Table 2). In addition, t _max was further delayed compared with Treatment A with a median difference (90% CI) of 1.00 h (0.50–1.75).

Midazolam and 1‐OH midazolam

Exposure to midazolam was higher during concomitant administration of 100 mg b.i.d. nivasorexant on Day 8 compared with midazolam alone (see Figure 3E and Table 2). GMRs (90% CI) between Treatment B and Treatment A of C _max, AUC_0–24, and t _½ were 1.34 (1.22–1.46), 1.56 (1.43–1.70), and 1.67 (1.38–2.01), respectively. For 1‐OH midazolam, GMRs of C _max and AUC_0–24 were unchanged while t _½ was slightly increased (1.29 [1.03–1.62], see Figure 3F and Table 2). Exposure to midazolam was even higher during Treatment D compared with Treatment A (see Figure 3E and Table 2). GMRs (90% CI) of C _max, AUC_0–24, and t _½ between both treatments were 2.62 (2.31–2.97), 3.71 (3.20–4.29), and 1.25 (1.07–1.46), respectively, while for 1‐OH midazolam, the values were 0.86 (0.75–0.98), 1.15 (1.06–1.24), and 3.05 (2.03–4.58), respectively (see Figure 3F and Table 2). Median t _max did not change for midazolam or its metabolite in any of the respective treatments.

3.2.3. Safety and tolerability

Administration of study treatments from Day 1 to Day 15 was safe and well tolerated. No serious AEs were reported. One subject discontinued study treatment due to an AE not related to any of the study treatments administered, that is, routine testing on Day 14 identified one subject as having contracted SARS‐CoV‐2 infection. The most commonly reported AE was somnolence, which was reported in Treatments B, C, and D by 1 (4.5%), 2 (9.1%), and 5 (23.8%) subject(s), respectively. All reported occurrences of somnolence were considered by the investigator to be related to study treatment. Other AEs reported on more than one occasion included fatigue, headache, nausea, and decreased appetite. No clinically relevant findings or changes from baseline in clinical laboratory, vital signs, ECG variables, or physical examination were observed in the study.

4. DISCUSSION

For the in vitro work, the effect of nivasorexant on the inhibition of CYPs was tested using human liver microsomes and CYP isoform‐specific marker transformations, or recombinant CYP2C19 expressed in baculovirus‐infected Sf9 cells. Thereby, IC₅₀ values in excess of 100 μM were determined for CYP1A2, CYP2A6, CYP2B6, and CYP2D6, with CYP3A4 exhibiting weak inhibition (IC₅₀ values of 19–44 μM). More pronounced inhibition was observed on all members of the CYP2C family, with IC₅₀ values in the range of 1.6–25 μM (see Table 1). In the ensuing clinical study, the inhibitory potential of nivasorexant on the two CYPs that showed the most pronounced inhibition in vitro (i.e., CYP2C9 and CYP2C19, with the addition of CYP3A4) was evaluated. A lower dose than administered in a previously conducted DDI study, that is, 100 mg b.i.d., was administered to match the dosing regimen of the Phase 2 POC study which was conducted in parallel. ⁹ Therefore, the effect of initial and repeated administration of the SO1RA nivasorexant on the PK of flurbiprofen, omeprazole, midazolam, and their CYP isoenzyme‐specific hydroxymetabolites in healthy subjects was assessed in a single‐center, open‐label, four‐period, fixed‐sequence Phase 1 study. The study design to simultaneously evaluate a potential DDI by making use of a cocktail approach was in line with current regulatory guidelines. ¹⁴ , ¹⁵ Omeprazole and midazolam are model substrates of CYP2C19 and CYP3A4, respectively, while flurbiprofen, a sensitive probe substrate was used to assess CYP2C9. ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ Several drugs including flurbiprofen, losartan, phenytoin, tolbutamide, and warfarin have been used as in vivo CYP2C9 probe drugs. ²⁰ , ²¹ , ²² Advantages of the use of flurbiprofen as a CYP2C9 probe substrate are its shorter t _½ compared with tolbutamide (~8 h) and warfarin (~36 h) which enables shorter PK sampling, while its PD effects provide less of a safety risk than tolbutamide or warfarin. ²² , ²³

Genotyping performed on Day 1 of the study showed the prevalence of determined alleles for CYP2C9, CYP2C19, and CYP3A4 to be in agreement with what has previously been reported. ²¹ , ²⁴ , ²⁵ , ²⁶ , ²⁷ Herein, all identified genotypes for CYP2C9, CYP2C19, and CYP3A4 are included in the assessment of the PK properties of the respective probe substrates with the exception of two poor metabolizers (i.e., one subject with CYP2C9 *2/*3 and another with CYP3A4 *22/*22), who were excluded from the analysis of flurbiprofen and midazolam, respectively. Although there was no significant difference between the analyses conducted with or without the inclusion of the respective subjects with a poor metabolizer genotype, the subjects were, nonetheless, excluded from analysis in accordance with regulatory guidance. ²⁸ Previously, it was determined both in vitro and in a clinical setting, that CYP3A4 *22 is associated with reduced hepatic CYP3A4 mRNA expression and CYP3A4 activity, respectively. ²⁹ , ³⁰ , ³¹ This has also been shown for CYP2C9 in subjects expressing *2/*3 genotype. ²¹ Herein, the plasma concentration–time profiles and PK characteristics of both subjects with the respective poor metabolizer genotype differed from the overall population that was included in the analysis of the effect of nivasorexant on the respective probe substrates (see Supplemental Figure S1, Supplemental Table S1), thereby, being in agreement with the aforementioned studies.

PK parameters of nivasorexant after single‐dose (on Day 8) as well as after repeated b.i.d. dosing (on Day 15) were comparable to PK parameters determined in a previous Phase 1 study wherein also 100 mg b.i.d. was administered. ⁸ After b.i.d. administration of 100 mg nivasorexant, mean trough nivasorexant concentrations reached steady state after 3–4 days' treatment, which is in line with the previous DDI study wherein steady‐state conditions were reached after 4–5 days' treatment with 200 mg b.i.d. nivasorexant. ⁹ An increase in AUC_τ after repeated b.i.d. dosing showed a 1.8‐fold accumulation of nivasorexant, also in line with previous findings. ⁸

When administered without nivasorexant, baseline flurbiprofen, omeprazole, and midazolam PK parameters were comparable to those reported previously under fasted conditions. ⁹ , ²¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷

According to regulatory guidance, an investigational drug can be classified as a weak, moderate, or strong inhibitor based on its effect on an appropriately chosen CYP substrate, whereby a weak, moderate, or strong inhibitor increases the AUC of the CYP substrate by ≥1.25‐ to <2‐fold, ≥ 2‐ to <5‐fold, or ≥5‐fold, respectively. ¹⁴ , ¹⁵ Accordingly, nivasorexant would be classified as a moderate inhibitor of CYP2C19 and a weak inhibitor of CYP3A4 after 1 day of 100 mg b.i.d. treatment and as a weak inhibitor of CYP2C9, a strong inhibitor of CYP2C19, and a moderate inhibitor of CYP3A4 after repeated b.i.d. administration of 100 mg, that is, at steady‐state conditions of nivasorexant. A lack of perpetrator DDI potential with respect to CYP2C9 was determined following initial administration of nivasorexant on Day 8 given that the 90% CIs of the GMR were all contained in the 0.8–1.25 bioequivalence boundaries and median t _max of flurbiprofen and 4‐OH flurbiprofen were not meaningfully different in the absence or presence of nivasorexant. This differed on Day 15, whereby an inhibitory effect of nivasorexant on flurbiprofen PK was observed, which was, however, to a lesser degree than has been reported previously for fluconazole, an index inhibitor of CYP2C9. Therein, an increase in AUC ratio of approximately two‐fold was observed when administered with the CYP2C9 probe substrates flurbiprofen, tolbutamide, and warfarin. ³⁸ , ³⁹ , ⁴⁰ Both initial as well as administration of nivasorexant under steady‐state conditions showed a pronounced effect on CYP2C19‐dependent omeprazole metabolism, which has previously been observed for strong CYP2C19 inhibitors, for example, fluvoxamine (25 mg b.i.d. for 6 days) which showed a comparable 5.6‐fold increase in omeprazole exposure. ⁴¹ The observed effect of nivasorexant on midazolam PK mirrors the previous findings that showed nivasorexant (i.e., after b.i.d. administration of 200 mg for 10 days) to be a moderate inhibitor of CYP3A4, which is similar to the reported effect following administration of both diltiazem (60 mg three times daily for 2 days) and fluconazole (400‐mg loading dose, followed by 200 mg per day for 3 days) on the PK of midazolam. ⁹ , ⁴² , ⁴³ The increase in midazolam AUC following initial and repeated b.i.d. administration of nivasorexant was in the same range in this study, albeit slightly lower than reported in the former study (i.e., midazolam AUC_0‐24 ratios in the former and present studies were 1.6 vs. 1.8 after first day of b.i.d. administration and 3.7 vs. 4.5 after repeated b.i.d. administration of nivasorexant, respectively). ⁹

The inhibitory effects of nivasorexant are not explained by competitive inhibition of the various CYP enzymes. Total C _max values on Days 8 and 15 were 693 ng/mL and 1252 ng/mL, respectively, corresponding to 16 and 29 nM unbound peak concentration after correction for molecular weight (429.5 g/mol) and plasma protein binding (f _u,p 0.01). These unbound C _max are significantly below the IC₅₀ values on competitive CYP inhibition (Table 1), thereby ruling out a direct inhibitory effect of the parent drug. Therefore, the strong pre‐incubation effect of nivasorexant on CYP2C19 observed in vitro, points towards the presence of inhibitory metabolites but does not allow for a conclusion on the exact inhibitory mechanism. Formation of metabolites with strong competitive inhibitory potential is one possibility, formation of a reactive metabolite that inactivates CYP2C19 by covalent binding is another. The latter hypothesis is supported by the observation that nivasorexant covalently binds to human liver microsomes and recombinant CYP2C19 (data not shown).

The mechanism for the effects on CYP2C9 and CYP3A4 is less clear as the corresponding in vitro data indicate no pre‐incubation effect on either CYP enzyme – at least not under the experimental conditions required by FDA and EMA regulatory guidelines. ¹⁴ , ¹⁵ , ²⁸ The observed covalent binding to both CYP enzymes indicates that reactive metabolites might also be involved in the inhibition of both CYP2C9 and CYP3A4. More detailed mechanistic work is being performed to elucidate the underlying mechanisms in detail.

Overall, treatments administered in this study were safe and well tolerated. The most frequently reported AEs were somnolence, headache, and fatigue which is in agreement with results from previous studies. ¹ , ⁸ , ⁹ In the current study, somnolence was recorded more frequently following concomitant administration of the probe substrates with nivasorexant at steady state (on Day 15) than when initially administered on Day 8. The observed higher incidence of somnolence is thought to be derived from the general increased exposure to study treatments, that is, midazolam and nivasorexant, on Day 15 compared with Day 8.

5. CONCLUSIONS

Compared with the other investigated CYPs in vitro, nivasorexant exhibited a more pronounced inhibition on CYP2C19 and CYP2C9, with IC₅₀ values of 1.6 and 8.6 μM, respectively, whereby nivasorexant was further shown to be a time‐dependent inhibitor of CYP2C19 only. Furthermore, CYP3A4 was inhibited with IC₅₀ values of 19–44 μM. The clinical study, which followed a cocktail approach, showed that treatment with flurbiprofen, omeprazole, midazolam, nivasorexant, and the combination thereof was safe and well tolerated based on AEs and other safety data. An increased exposure to omeprazole and midazolam of 2.1‐ and 1.6‐fold, respectively, was observed upon initial administration of 100 mg b.i.d. nivasorexant. Furthermore, an increased exposure to flurbiprofen, omeprazole, and midazolam of 1.5‐, 6.8‐, and 3.7‐fold, respectively, was observed after concomitant administration of 100 mg b.i.d. nivasorexant under steady‐state conditions. According to regulatory guidance, nivasorexant is classified as a moderate inhibitor of CYP2C19 and weak inhibitor of CYP3A4 after 1 day of 100 mg b.i.d. treatment and as a weak inhibitor of CYP2C9, a strong inhibitor of CYP2C19, and a moderate inhibitor of CYP3A4 after repeated b.i.d. administration of 100 mg, that is, at steady state. ¹⁴ , ¹⁵ The in vitro results suggest that inhibitory or reactive metabolites of nivasorexant (possibly in conjunction with covalent binding) could be acting as perpetrator(s). More detailed in vitro work is required to trace the exact underlying workings. Nevertheless, it can be assumed, that the observed effects of nivasorexant in healthy subjects would occur with other medications that are regulated or metabolized by the affected CYPs investigated herein.

AUTHOR CONTRIBUTIONS

AT and NG performed preclinical in vitro experiments; BB, PK, and JD designed and supervised the clinical study; MB conducted the clinical part; BB, PK, AT, NG, and JD analyzed the respective data; BB wrote the manuscript; All authors reviewed and approved the final manuscript. The authors confirm that the principal investigator for this study was MB and that he had direct clinical responsibility for the subjects.

FUNDING INFORMATION

This study was funded by Idorsia Pharmaceuticals Ltd.

CONFLICT OF INTEREST STATEMENT

BB, PK, AT, NG, and JD are employees of Idorsia Pharmaceuticals Ltd and own stocks (options) of Idorsia Pharmaceuticals Ltd. MB was the principal investigator and an employee of CRS Berlin GmbH. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the article.

ETHICS STATEMENT

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

INFORMED CONSENT

Informed consent was obtained from all individual participants included in the study.

Supporting information

Data S1:

Figure S1.

Berger B, Kaufmann P, Berse M, Treiber A, Grignaschi N, Dingemanse J. Effect of nivasorexant (ACT‐539313), a selective orexin‐1‐receptor antagonist, on multiple cytochrome P450 probe substrates in vitro and in vivo using a cocktail approach in healthy subjects. Pharmacol Res Perspect. 2023;11:e01143. doi: 10.1002/prp2.1143

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.