Respiratory and Antinociceptive Effects of Nociceptin Receptor–µ-Opioid Receptor Agonist Cebranopadol versus Full Opioid Receptor Agonist Oxycodone: A Comparison in Healthy Volunteers

Abstract

Background:

The novel analgesic cebranopadol targets the nociceptin (NOP) and µ-opioid (MOP) receptors, acting as a novel full dual NOP–MOP receptor agonist, with possible differences in respiratory effects compared to selective MOP opioids like oxycodone.

Methods:

In this randomized, double-blind, placebo-controlled study, 30 healthy volunteers received oral placebo (n = 20), cebranopadol (600 µg, n = 20; 800 µg, n = 20; or 1,000 µg, n = 20), or oxycodone (30 mg, n = 20; or 60 mg, n = 20) on four occasions in a partial crossover design. On each occasion, measures of ventilation at an extrapolated isohypercapnic level of 55 mmHg derived from hypercapnic ventilatory responses and electrical pain tolerance tests were obtained at regular intervals before and for 24 h after drug intake. Mixed model analyses on respiratory endpoints were performed (primary endpoint) as well as an exploratory population pharmacokinetic/pharmacodynamic analysis on respiratory and analgesic endpoints.

Results:

Oxygen desaturations (to approximately 80%) were observed in 65% of subjects after oxycodone 60 mg versus cebranopadol 1,000 µg in 25% of subjects (all occurring when experimental respiratory or pain testing was not performed). A significant main effect and a significant separation of all cebranopadol and oxycodone doses versus placebo (all P < 0.0001) was observed with cebranopadol 600 µg producing less respiratory depression than oxycodone 30 mg (P = 0.022). Pharmacokinetic/pharmacodynamic analyses showed that the respiratory drug concentration causing 50% effect was 0.20 ± 0.54 ng/ml for cebranopadol versus 36 ± 6 ng/ml for oxycodone (values are median ± standard error of the estimate). Cebranopadol was more potent than oxycodone in producing analgesia.

Conclusions:

The primary endpoint showed separation between the respiratory effects of cebranopadol and oxycodone, with 25% less respiratory depression at equianalgesia, as observed in the pharmacokinetic/pharmacodynamic analysis.

This study found that cebranopadol induced less respiratory depression than oxycodone at equivalent analgesic doses and resulted in fewer and less severe oxygen desaturation events. Cebranopadol was also more potent, had a slower onset and longer duration of effect, and showed a promising safety profile, supporting its potential as a safer alternative for pain management; however, further research is needed.

Editor’s Perspective.

What We Already Know about This Topic

Opioids like oxycodone are effective for moderate to severe pain but carry high risks of respiratory depression and abuse. Cebranopadol has shown promise as a potent analgesic with fewer adverse effects, including reduced respiratory depression, compared to traditional opioids due to its dual effects on µ-opioid receptors and nociceptin receptors. However, these safety advantages had not been fully confirmed in direct, high-dose human comparisons.

What This Article Tells Us That Is New

This study found that cebranopadol induced less respiratory depression than oxycodone at equivalent analgesic doses and resulted in fewer and less severe oxygen desaturation events. Cebranopadol was also more potent, had a slower onset and longer duration of effect, and showed a promising safety profile, supporting its potential as a safer alternative for pain management; however, further research is needed.

Opioids acting at the µ-opioid peptide (MOP) receptor remain the only effective option for treating moderate to severe pain in many patients.¹ However, their use is tempered by safety concerns, including the risks of dependence, misuse, and respiratory depression.^1,2 Particularly respiratory depression may have significant sequelae, even in nonfatal cases.^1,3 Therefore, there is an unmet need for a therapy that provides analgesia comparable to opioids with an improved safety profile.

In 1994, the nociceptin/orphanin FQ peptide (NOP) receptor, a G protein–coupled receptor with approximately 65% structural homology with opioid receptors but with virtually no affinity for opioid ligands, was identified.^4–6 This receptor is widely distributed throughout the central nervous system.⁷ Simultaneous agonism of the NOP and MOP receptors has been shown to produce additive or synergistic analgesic effects. Importantly, NOP agonism counteracts several negative effects of MOP agonism such as respiratory depression and abuse potential.^8,9 This allowed for the development of an analgesic that provides relief for moderate to severe pain comparable to that of pure MOP agonists, but with an improved safety profile.^10–12

Cebranopadol was designed specifically to target the NOP and MOP receptors in roughly equal amounts, acting as a novel full dual NOP–MOP receptor agonist.^13,14 To date, several nonclinical and clinical studies have demonstrated that it provides opioid-level analgesia in nociceptive and neuropathic pain with lesser adverse effects, including lower abuse potential.^15–19 In addition, several nonclinical and pilot clinical studies showed that cebranopadol produced minimal respiratory depression at analgesic or supra-analgesic doses.^8,15,20,21

This lack of respiratory depression is due to cebranopadol’s NOP agonism.²⁰ A study in healthy human volunteers who received a single 600-µg dose of cebranopadol demonstrated a slow decrease in ventilation developed under the experimental conditions of isohypercapnia with negligible changes in oxygen saturation.²¹ Data from the animal and human studies suggest that cebranopadol may have an apparent plateau in respiratory effects above concentrations produced by the 600-µg dose studied.^15,20,21

The primary objective of this study was to compare the respiratory effects of cebranopadol to the commonly used MOP agonist, oxycodone; additionally, we constructed a pharmacokinetic (PK)/pharmacodynamic (PD) model. We hypothesize that cebranopadol produces less respiratory depression than oxycodone at equianalgesia.

Materials and Methods

This was a randomized, double-blind, four-period, six-treatment, placebo-controlled, partial crossover study to investigate the hypercapnic ventilatory response (HCVR),²² nociceptive thresholds, pharmacokinetics, and safety of oral cebranopadol and oxycodone in healthy adults. Data on all endpoints (electrical pain threshold, HCVR, pupil diameter, opioid plasma concentrations, safety data) were collected at several fixed time points and continuously monitored during a 24-h period with respect to safety parameters. We first performed a linear mixed model (LMM) analysis of the respiratory data (primary endpoint) with the main focus on the main treatment effects. Separately (secondary analysis), we created a PK/PD model to characterize the drug effects of cebranopadol and the active comparator, oxycodone, using the following data: respiration, pain threshold, and pupil diameter. While the LMM analysis relates dose to effect, the PK/PD analyses relate drug plasma concentration to effect. Both model the measured responses, albeit with distinct models, and both give valuable information on a possible separation between cebranopadol and oxycodone.

We realized in retrospect that our study design may make separation of effect difficult for several reasons: (1) the drugs differ substantially in PK profiles such that cebranopadol is expected to have a slower onset of action and much longer duration of effect; (2) peak respiratory effects might not occur at a specific timepoint measured; (3) substantial variability was observed in each of the individual endpoints collected because of high intersubject variance; (4) we applied a partial crossover design, and as a main consequence, not every subject received every treatment. This may result in overlapping CIs and may limit the power of the LMM. In addition, the oxycodone and cebranopadol doses did not produce equivalent analgesia, further limiting the interpretability of the LMM. A model of effect versus plasma concentration overcomes most of these limitations.

Ethics and Subjects

Ethics

The study was approved by the Medical Ethics Committee Leiden, The Hague, Delft (The Netherlands; www.metc-ldd.nl; date of approval July 15, 2022), and preregistered at the International Standard Randomised Controlled Trial Number registry (www.isrctn.com) under No. ISRCTN11320277 (www.isrctn.com/ISRCTN11320277) on July 27, 2022 (main contact, Simone Jansen, M.D.). The study was performed from July 29, 2022, until April 25, 2023, at the Center for Human Drug Research (Leiden, The Netherlands) and at Leiden University Medical Center (LUMC; Leiden, The Netherlands) with principal investigator Albert Dahan, M.D., Ph.D. No changes were made to the protocol or trial outcomes after trial registration. All subjects gave written informed consent before participation in the trial.

Subjects

Healthy volunteers of either sex were recruited to participate in the trial. Details of the inclusion and exclusion criteria are given in the Supplemental Digital Content (https://links.lww.com/ALN/E330). In brief, inclusion criteria were age 18 to 45 yr (inclusive), body mass index 18 to 32 kg/m² (inclusive), and body weight 50 kg or more; exclusion criteria were history or presence of clinically relevant medical, neurologic, or psychiatric disease; abnormalities on the electrocardiogram; systolic blood pressure greater than 150 mmHg or less than 90 mmHg; diastolic blood pressure greater than 95 mmHg or less than 50 mmHg; heart rate less than 40/min or greater than 100/min during rest; a positive alcohol test or urine test for drugs of abuse; pregnancy or breastfeeding; any allergic history; and use of any medication (with the exception of contraceptives and incidental use of paracetamol or ibuprofen, which were allowed up to 48 h before start of each visit) or herbal preparations containing St. John’s wort. The subjects were admitted to the clinical research unit of the Center for Human Drug Research for one night and the Anesthesia & Pain Research Unit at LUMC on the day of the study for the experimental tests, where they stayed overnight.

Study Design

Treatment, Blinding, Randomization, and Dispensing

The following oral treatments were tested: placebo (LUMC pharmacy for oxycodone, and Patheon France SAS, France, for cebranopadol), cebranopadol 600, 800, or 1,000 µg (Patheon France SAS), and oxycodone 30 or 60 mg (TEVA Nederland BV, The Netherlands). The drugs were overencapsulated by the pharmacy to ensure complete blinding. The treatments were ingested with 240 ml noncarbonated water.

Enrolled subjects were randomized to a unique administration sequence of four (of six possible) treatments that was administered on more than four visits (periods) to the research unit, with each period separated by a 14-day washout period (fig. 1A). The treatment sequence for each of the participants is given in supplemental table 1 (https://links.lww.com/ALN/E330). On each treatment day, the HCVR, nociceptive thresholds, and pupil diameter were tested at regular intervals. A final safety follow-up visit was conducted 14 days after the last treatment was received. Subjects were randomized to study treatment in a consecutive order starting with the lowest number according to a randomization code generated by a study-independent statistician at the study center. The randomization code was shared with the pharmacy but otherwise kept strictly confidential. The pharmacy prepared the medication and dispensed the treatment on the morning of the study in unmarked containers. All capsules were identical in size, taste, and smell.

Fig. 1.

(A) Study protocol. There were 1 to 2 weeks in between treatment days. (B) Consolidated Standards of Reporting Trials (CONSORT) flow diagram.

Dose Selection

The expected dose range for cebranopadol for the treatment of pain is 200 to 400 µg per day.^11,18,19,22 We selected the three supra-analgesic doses of 600 µg, 800 µg, and 1,000 µg for this trial in order to test doses as high as possible to simulate a worst-case scenario for respiratory effects and directly test for an apparent plateau in respiratory depression. It was previously demonstrated that most healthy volunteers cannot tolerate single oral doses of 800 µg or more without vomiting in the absence of a titration period.^18,19 However, we aimed to study doses in this range to fully characterize the respiratory safety of the drug and because there is variability in tolerability. To minimize these side effects in the trial and allow for doses at these levels, each subject was dosed with 4 mg intravenous ondansetron prophylactically.

Oxycodone doses of 30 mg and 60 mg were selected because they were expected to produce significant respiratory depression and expected to produce analgesia similar to the cebranopadol doses based on previous efficacy studies.^23–25 However, the relative antinociceptive potency of the drugs was not precisely known and was one of endpoints measured in this trial.

Respiratory Testing

In order to obtain an indication of isohypercapnic ventilation at an end-tidal carbon dioxide concentration (PCO₂) of 55 mmHg (${\dot{\mathrm{V}}}_{\mathrm{E}}55$), we obtained modified Read carbon dioxide rebreathing responses at 12 timepoints: twice before dosing and at 1, 2, 3, 4, 5, 6, 7, 8, 10, and 24 h after dosing.^23–25 Subjects were comfortably seated in a semirecumbent position. During respiratory testing, they wore a facemask connected to a pneumotachograph and pressure transducer system (Hans Rudolph Inc., USA) to measure breath-to-breath minute ventilation. Additionally, we measured inspired and expired gas concentrations at the mouth using a Masimo Root ISA OR plus capnograph (Masimo, USA). At each respiratory testing period, the subjects initially breathed room air for 4 min, after which they were coached to hyperventilate for 2 to 3 min while breathing 100% oxygen. Thereafter they were connected to a 6-l rebreathing bag containing 7% carbon dioxide and 93% oxygen.^23–25 Subjects continued rebreathing for 3 to 4 min. The complete test lasted about 10 min. In between tests, the facemask was removed, and the subjects breathed ambient air. Linear regression was performed in R (available at www.R-project.org/, accessed January 14, 2026) on the carbon dioxide–dependent part of the ${\dot{\mathrm{V}}}_{\mathrm{E}}$–carbon dioxide response curve (black dots in fig. 2) according to the equation ${\dot{\mathrm{V}}}_{\mathrm{E}}$ = S × (end-tidal PCO₂ – B), where S is the slope of the response curve and B is the apneic threshold or extrapolated end-tidal PCO₂ at which apnea occurs theoretically.^23–25 See figure 2 for an example of the HCVR and the regression analysis. The initial flat part of the curve (gray dots) is the carbon dioxide–independent part of the response curve and not included in the regression analysis. By inputting 55 for end-tidal PCO₂ in the equation, we obtain the ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ value (fig. 2). This isohypercapnic value was used in the data analysis as it is considered one of the most sensitive tests of opioid effect on ventilatory control.^22,23

Fig. 2.

Hypercapnic ventilatory response obtained by the rebreathing technique in a single subject 1 h after oxycodone 30 mg administration. Each dot is one breath. The curve has two parts: a horizontal part, where ventilation is independent of carbon dioxide (the wakefulness drive), depicted by gray symbols, and a linear increasing part, i.e., where ventilation increases with increasing carbon dioxide levels in the end-expiratory gas (black dots). The carbon dioxide–dependent data (black dots) are analyzed by linear regression through the linear part of the response data according to the formula ${\dot{\mathrm{V}}}_{\mathrm{E}}$ = S × (end-tidal carbon dioxide concentration [PCO₂] – B), where S is the slope of the response curve and B the apneic threshold or extrapolated end-tidal PCO₂ at which apnea occurs theoretically. By inputting 55 for end-tidal PCO₂, the ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ value is obtained (red arrows). In this example, the slope of the hypercapnic ventilatory response (S) = 11.5 l/min per kPa, the apneic thershold (B) = 5.8 kPa, and extarpolated ventilation at 55 mmHg carbon dioxide concentration (${\dot{\mathrm{V}}}_{\mathrm{E}}55$) = 17.3 l/min.

Antinociceptive Testing

In order to get an indication of the antinociceptive potency of the two test drugs, we obtained antinociceptive responses, including the tolerance threshold during transcutaneous electrical stimulation.²⁶ To that end, two surface electrodes (Red Dot, 3M, Canada) were placed on the skin overlying the tibial bone (shin bone) of the left leg. The electrodes were attached to a computer-interfaced constant current stimulator (Leiden University, The Netherlands). The intensity of the noxious stimulation was increased from 0 mA in steps of 0.5 mA/s with a pulse duration of 0.2 ms at 10 Hz and a cutoff of 128 mA. The subjects were instructed to press a button on a control box when no further increase in stimulus intensity was acceptable to them (pain tolerance). When the subject pressed the button, the stimulus train ended, and the current was collected and stored for further analysis. Antinociceptive responses were obtained twice predose and at 30 min, 1.5, 2.5, 3.5, 4.5, 5.5, 6.5, 7.5, 9.5, and 23.5 h after dosing.

Pupil Diameter

Pupillometry was conducted using the handheld Neuroptics PLR-3000 pupillometer (Neuroptics, USA).²⁷ The pupillometer was positioned against the orbital rim of either the right or left eye, creating a secure seal with the rubber part of the scanner, and proceeded to automatically detect and scan the pupil diameter. The pupil diameter was measured twice before the administration of the study drug and at specific time intervals: 30 min and 1.5, 2.5, 3.5, 4.5, 5.5, 6.5, 7.5, 9.5, and 23.5 h after dosing, all performed before the subsequently planned pain test.

Blood Sampling

Venous blood samples of 6 ml were drawn on study days at t = 0 (predosing) and at t = 30 min and 1, 1.5, 2, 3, 4, 5, 7, 8, 9, and 24 h after dosing (total blood volume drawn on the 4 study days = 288 ml). The samples were collected in prechilled sodium fluoride/potassium oxalate tubes and then centrifuged at 4°C for 10 min at 3,000g within 45 min of collection. This process yielded approximately 2.5 ml plasma, which was split into two separate aliquot tubes and stored at −80°C until shipment. The samples were later analyzed at Pharma Medical Research, Inc., in Mississauga, Canada, using validated Liquid Chromatography-Tandem Mass Spectrometry methods for oxycodone and cebranopadol concentrations.

Safety

Throughout the study day (24 h), we measured vital signs (oxygen saturation, blood pressure, heart rate). If subjects experienced severe respiratory events unresponsive to basic measures such as verbal or tactile stimulation and/or supplemental oxygen, intravenous naloxone 0.4 mg could be administered. Such events included oxygen desaturation less than 90% for more than 1 min, more than three events per minute unresponsive to conservative measures, apnea more than 20 s, and oxygen desaturation less than 90% more than two times per minute despite conservative measures.

Data Analysis

Sample Size

The sample size of 30 subjects that completed the study was based on a previously completed study that also evaluated cebranopadol in healthy volunteers. The six-way crossover design with four treatments per subject results in complete crossover contrasts of 12 subjects. Supplemental table 1 (https://links.lww.com/ALN/E330) gives the treatment sequences for each of the 30 participants. A power calculation using a sample size of 12 revealed a power of 0.80 to detect a difference in means of 1.29 l/min (minute ventilation), assuming an SD of differences of 1.45 l/min, and using a paired t test with a 0.05 two-sided significance level (SAS 9.4; SAS Institute, USA).

Statistical Analysis

To establish whether significant treatment effects could be detected on the primary endpoint (i.e., ${\dot{\mathrm{V}}}_{\mathrm{E}}55$), it was analyzed with an LMM analysis of covariance with treatment, time, period, and treatment by time as fixed factors and subject, subject by treatment, and subject by time as random factors and the average baseline measurement as a covariate. The data were initially analyzed without transformation, but if the data suggested otherwise, log-transformation was applied. The general treatment effect and specific contrasts for the primary endpoint (${\dot{\mathrm{V}}}_{\mathrm{E}}55$) were reported with the estimated difference and the 95% CI, the least square mean estimates, and the P value. The analysis of ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ was performed in sequence to limit the number of tests (i.e., hierarchical testing). First, only a treatment effect was generated to determine whether there was a significant (P < 0.05) difference between treatments. Next, where relevant, contrasts between individual treatments were calculated within the model. Other PD endpoints were presented in graphs.

PK/PD Analyses

Population PK models were developed separately for oxycodone and cebranopadol as specified in Supplemental Digital Content paragraph 3 (https://links.lww.com/ALN/E330). In brief, for both cebranopadol and oxycodone PK data, one and two compartment models were tested with absorption compartments and with consideration of lag time.

Population PD sigmoidal maximum effect models were constructed separately for each of the endpoints and test drugs. The analyses are specified in Supplemental Digital Content paragraph 4 (https://links.lww.com/ALN/E330). In brief, for each endpoint, blood concentrations to be used in the PD analysis were obtained from the PK analysis, individualized using the empirical Bayesian estimates of the model parameters. Important model parameters that were estimated are the blood–effect site equilibration half-life (t½k_e0), which identifies the hysteresis between plasma drug concentration and effect; the potency parameter (concentration at which 50% of maximum effect occurs [C₅₀]), which identifies the strength of the drug causing a 50% change in effect relative to the maximum effect; maximum PD effect; and shape parameter γ, which represents the steepness of the concentration–effect sigmoidal response.

To determine the performance of the PK and PD models, goodness-of-fits plots and visual predictive checks were plotted to assess the adequacy of the models. Data analyses were conducted in Nonlinear Mixed Effects Modeling (NONMEM) 7.5.1 (Icon Plc., USA) with P values < 0.05 considered significant.

Finally, to get a better insight into our respiratory results at equianalgesia, we obtained 10,000 samples from the distributions of for ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ and pain tolerance in NONMEM. The equianalgesic ratio for the oxycodone versus cebranopadol was assessed for each set of samples of the tolerance C₅₀ values and next normalized ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ values were calculated for each set of samples of the ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ C₅₀ values. Next, 90% CIs were determined of these ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ values and their differences.

Results

Fifty-three subjects were screened for eligibility, of whom 23 were excluded for various reasons (logistic, exclusion criteria; see the Consolidated Standards of Reporting Trials [CONSORT] flow diagram in fig. 1B). Thirty subjects were randomized, and they completed all four visits without serious side effects; no subjects withdrew from the study. The characteristics of the 30 subjects (18 men and 12 women) were as follows: age, 27.7 ± 5.8 yr; body mass index, 23.1 ± 2.6 kg/m²; height, 175 ± 8 cm; and weight, 71 ± 13 kg. Twenty subjects were White, three Asian, two Black, and four of mixed origin; the origin of one subject remained unspecified.

Safety

All subjects completed the study without serious adverse events. All but one subject who received 600 and 800 µg cebranopadol and 60 mg oxycodone experienced at least one adverse event; all subjects treated with 1,000 µg cebranopadol and 30 mg oxycodone reported adverse events (see table 1 for the adverse events collected according to the Medical Dictionary for Regulatory Activities [MedDRA] system). A general trend of increasing total adverse event count with increasing dose was noted, consistent across both cebranopadol and oxycodone treatment groups. The majority of adverse events (95%) were mild in severity. The most common adverse events for cebranopadol were nausea, somnolence, and dizziness, while for oxycodone, adverse events were somnolence, decreased oxygen saturation, and nausea. Desaturations (oxygen saturation decreasing to approximately 80%) occurred particularly after administration of 60 mg oxycodone, with 65% of subjects having one or more desaturation events as compared to 25% after cebranopadol 1,000 µg (the highest rate for any of the cebranopadol doses), which were related to short episodes of apnea (approximately 10 s) occurring in between respiratory testing and resolving without medical intervention. Stimulation of the subjects to take a deep breath was sufficient to normalize oxygen saturation. No adverse events necessitated discontinuation of study treatment, withdrawal from the study, or use of naloxone.

Table 1.

Adverse Events by Medical Dictionary for Regulatory Activities (MedDRA) System Organ Class (SOC) and Preferred Term

	No. of Subjects (%)
SOC	Placebo (n = 20)	Cebranopadol 600 µg (n = 20)	Cebranopadol 800 µg (n = 20)	Cebranopadol 1,000 µg (n = 20)	Oxycodone 30 mg (n = 20)	Oxycodone 60 mg (n = 20)
Gastrointestinal disorders
Constipation	—	—	1 (5.0%)	2 (10.0%)	2 (10.0%)	—
Dry mouth	—	1 (5.0%)	—	3 (15.0%)	1 (5.0%)	3 (15.0%)
Nausea	1 (5.0%)	11 (55.0%)	10 (50.0%)	14 (70.0%)	12 (60.0%)	12 (60.0%)
Vomiting	—	6 (30.0%)	6 (30.0%)	8 (40.0%)	4 (20.0%)	7 (35.0%)
Respiratory, thoracic, and mediastinal disorders
Apnea	—	2 (10.0%)	1 (5.0%)	2 (10.0%)	1 (5.0%)	4 (20.0%)
Hiccups	—	2 (10.0%)	1 (5.0%)	4 (20.0%)	2 (10.0%)	3 (15.0%)
Respiratory depression	—	—	1 (5.0%)	1 (5.0%)	1 (5.0%)	3 (15.0%)
Oxygen desaturation	1 (5.0%)	3 (15.0%)	4 (20.0%)	5 (25.0%)	2 (10.0%)	13 (65.0%)
Nervous system disorders
Dizziness	2 (10.0%)	10 (50.0%)	8 (40.0%)	8 (40.0%)	9 (45.0%)	8 (40.0%)
Headache	2 (10.0%)	7 (35.0%)	3 (15.0%)	3 (15.0%)	8 (40.0%)	3 (15.0%)
Somnolence	3 (15.0%)	10 (50.0%)	12 (60.0%)	10 (50.0%)	14 (70.0%)	14 (70.0%)
General disorders and administration site conditions
Chills	—	1 (5.0%)	1 (5.0%)	3 (15.0%)	—	—
Discomfort	—	2 (10.0%)	1 (5.0%)	1 (5.0%)	—	1 (5.0%)
Fatigue	1 (5.0%)	5 (25.0%)	1 (5.0%)	5 (25.0%)	3 (15.0%)	1 (5.0%)
Feeling hot	—	—	—	—	—	3 (15.0%)
Feeling of relaxation	—	1 (5.0%)	—	—	2 (10.0%)	1 (5.0%)
Ear and labyrinth disorders
Vertigo	—	1 (5.0%)	—	—	2 (10.0%)	—
Vascular disorders
Flushing	2 (10.0%)	1 (5.0%)	1 (5.0%)	1 (5.0%)	3 (15.0%)	2 (10.0%)
Skin and subcutaneous tissue disorders
Pruritus	—	1 (5.0%)	3 (15.0%)	4 (20.0%)	8 (40.0%)	10 (50.0%)
Psychiatric disorders
Euphoric mood	—	—	6 (30.0%)	—	10 (50.0%)	11 (55.0%)

Statistical Analysis of the Partial Crossover Trial

The results of the comparisons of ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ for the two drugs using the LMM analysis of covariance are given in table 2. The data show significant main effects and a significant separation of all cebranopadol and oxycodone doses versus placebo: treatment P < 0.0001, time P < 0.0001, and treatment by time P < 0.001. Despite our expectation that separation among doses would be difficult, we performed post hoc comparisons between the different doses of cebranopadol and oxycodone and observed that these did not reach the level of significance except for the comparison of cebranopadol 600 mg versus oxycodone 30 mg (P = 0.022). Interestingly, the magnitude of respiratory depression (${\dot{\mathrm{V}}}_{\mathrm{E}}55$) was lesser in the participants that received 60 mg oxycodone compared to those that received 30 mg oxycodone. Consequently, while the comparison of 600 µg cebranopadol with 30 mg oxycodone was significantly different (P = 0.022), this was not case for 600 µg cebranopadol with 60 mg oxycodone (P = 0.40). We relate this to the partial crossover design and insufficient overlap of subjects in the different treatment groups (particularly the 30 and 60 mg oxycodone groups).

Table 2.

Mixed Model Analysis of ${\dot{\mathrm{V}}}_{\mathrm{E}}55$

					95% CI
Effect/Contrast	P Value	First LSM of Contrast	Second LSM of Contrast	Estimate of the Difference	Lower	Upper
Prevalue	< 0.0001	—	—	—	—	—
Treatment	< 0.0001	—	—	—	—	—
Period	0.2159	—	—	—	—	—
Time	< 0.0001	—	—	—	—	—
Treatment by time	< 0.0001	—	—	—	—	—
Cebranopadol 1,000 µg - placebo	< 0.0001	17.2378	26.4647	−9.2269	−12.2230	−6.2307
Cebranopadol 800 µg - placebo	< 0.0001	18.4353	26.4647	−8.0294	−11.0340	−5.0248
Cebranopadol 600 µg - placebo	< 0.0001	18.9741	26.4647	−7.4906	−10.5067	−4.4746
Oxycodone 30 mg - placebo	< 0.0001	15.4491	26.4647	−11.0156	−14.0126	−8.0186
Oxycodone 60 mg - placebo	< 0.0001	17.6976	26.4647	−8.7671	−11.7639	−5.7703
Cebranopadol 1,000 µg - oxycodone 60 mg	0.7611	17.2378	17.6976	−0.4598	−3.4578	2.5382
Cebranopadol 800 µg - oxycodone 60 mg	0.6277	18.4353	17.6976	0.7377	−2.2755	3.7509
Cebranopadol 800 µg -oxycodone 30 mg	0.0512	18.4353	15.4491	2.9862	−0.0161	5.9886
Cebranopadol 600 µg - oxycodone 60 mg	0.4043	18.9741	17.6976	1.2764	−1.7517	4.3046
Cebranopadol 600 µg - oxycodone 30 mg	0.0223	18.9741	15.4491	3.5250	0.5135	6.5364

LSM, least significant difference; ${\dot{\mathrm{V}}}_{\mathrm{E}}55$, ventilation at 55 mmHg carbon dioxide.

PK Analyses

The average drug concentration profiles for cebranopadol and oxycodone for the complete 24-h sampling period are given in figure 3. Peak plasma concentrations for cebranopadol 600, 800, and 1,000 µg were 178.2 ± 41.2, 210.7 ± 41.2, and 227.2 ± 34.6 pg/ml, respectively, occurring at t = 6 h for all three doses. Peak plasma concentrations for oxycodone 30 mg and 60 mg were 42.3 ± 9.0 ng/ml and 68.9 ± 28.0 ng/ml, respectively, at t = 1.5 h for both doses. A description of PK model building and PK analyses is given in Supplemental Digital Content paragraph 3 (https://links.lww.com/ALN/E330). The final PK model is given in figure 4.

Fig. 3.

Cebranopadol (A) and oxycodone (B) plasma concentrations. Data are mean ± SD. Conc., concentration.

Fig. 4.

Pharmacokinetic model. V is the volume of a specific compartment with V_abs, an absorption compartment with absorption rate constant k_a. V_central is the central compartment, and V_periph is a proposed peripheral compartment that was not included in the final model. CL is clearance and Q the proposed intercompartmental clearance. Finally, t_lag is the intestinal absorption lag time.

PD Analyses

The analyses were conducted on ${\dot{\mathrm{V}}}_{\mathrm{E}}55$, electrical pain tolerance, and pupil diameter responses. In supplemental figure 4 (https://links.lww.com/ALN/E330), we give the mean data ± 95% CI of these endpoints; the different depicted doses are the mean of separate sets of subjects. We tested various assumptions regarding the PD models (Supplemental Digital Content paragraph 4.2.1, https://links.lww.com/ALN/E330). Table 3 presents the estimated PD parameters including between-subject and interoccasion variabilities and residual errors. The data fits, goodness-of-fit plots, and visual predictive checks indicate that the PD models for all three endpoints adequately described the data; see for the graphs supplemental figures 5A and 6C (https://links.lww.com/ALN/E330).

Table 3.

Pharmacodynamic Parameter Estimates for Cebranopadol and Oxycodone

Parameter	Estimate ± SEE	Intersubject Variability ± SEE	CV (%)	Interoccasion Variability ± SEE	CV (%)
Cebranopadol
${\dot{\mathrm{V}}}_{\mathrm{E}}55$
Baseline (l/min)	26.2 ± 1.9	0.1 ± 0.045	32	0.1 ± 0.033	32
C50 (pg/ml)	197 ± 54	182 ± 0.61	227
γ	1	0.59 ± 0.21	89
t½ke0 (min)	63 ± 16
σ2	22.5 ± 1.4
σe2	82 ± 20
Electrical pain tolerance
Baseline (mA)	19.4 ± 1.8	0.239 ± 0.061	52	0.0301 ± 0.0076	17
C50 (pg/ml)	1,270 ± 180			0.65 ± 0.32	95
γ	1			0.373 ± 0.085	67
t½ke0 (min)	128 ± 13
σ2	7.46 ± 0.31
σe2	11.2 ± 2.6
Miosis
Baseline (mm)	4.41 ± 0.11	0.0146 ± 0.0045	12	0.0038 ± 0.0013	6
C50 (pg/ml)	1,060 ± 230	0.64 ± 0.27	95
γ	0.519 ± 0.059
t½ke0 (min)	112 ± 45
σ2	0.117 ± 0.0071
σe2	0.074 ± 0.038
Oxycodone
${\dot{\mathrm{V}}}_{\mathrm{E}}55$
Baseline (l/min)	26.6 ± 1.8	0.42 ± 0.18	72		39
C50 (ng/ml)	36 ± 5.7
γ	1			1.25 ± 0.64	158
t½ke0 (min)	7.6 ± 1.6
σ2	29.1 ± 2.2
σe2	105 ± 32
Electrical pain tolerance
Baseline (mA)	19.3 ± 1.5	0.128 ± 0.046	37	0.034 ± 0.014	18
C50 (ng/ml)	619 ± 230			1.1 ± 0.33	141
γ	1	0.24 ± 0.11	51
t½ke0 (min)	3.8 ± 2
σ2	9.76 ± 0.49
σe2	9.7 ± 3.4
Miosis
Baseline (mm)	4.41 ± 0.12	0.0143 ± 0.0057	12	0.007 ± 0.0043	8
C50 (ng/ml)	97 ± 14			0.31 ± 0.15	59
γ	0.691 ± 0.063			0.15 ± 0.069	40
t½ke0 (min)
σ2	0.088 ± 0.0068
σe2	0.104 ± 0.05

γ, A shape parameter; σ², variance of the additive intraindividual error; σ_e², supplementary variance introduced for the 24-h measurements with respect to σ²; baseline, parameter value before drug administration; C₅₀, concentration at which 50% of maximum effect occurs; CV, coefficient of variation; SEE, standard error of the estimate; t½k_e0, blood-effect site equilibration half-life; ${\dot{\mathrm{V}}}_{\mathrm{E}}55$, extrapolated isohypercapnic minute ventilation at 55 mmHg end-tidal carbon dioxide concentration.

Relevant parameter estimates were the C₅₀ values for ${\dot{\mathrm{V}}}_{\mathrm{E}}55$: 0.197 ± 54 ng/ml for cebranopadol versus 36 ± 6 ng/ml oxycodone; for pain tolerance, these values were 1.27 ± 0.18 ng/ml for cebranopadol versus 619 ± 230 ng/ml for oxycodone, and for miosis, 1.06 ± 0.23 ng/ml for cebranopadol and 97 ± 14 ng/ml for oxycodone. Estimates of t½k_e0 ranged from 1 to 2 h for cebranopadol for the three measured endpoints (approximately 1 h for ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ and 2 h for pain tolerance and miosis), while the values for oxycodone were approximately 8 and 4 min for ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ and pain tolerance, respectively.

To assess the relative respiratory depression induced by cebranopadol and oxycodone, we used the PK and PD data to quantify their respiratory effects at increasing plasma concentrations in equianalgesic conditions. The results depicted in figure 5 demonstrate that, at comparable levels of analgesia, cebranopadol produces 10 to 25% less respiratory depression (P < 0.05). This effect is consistent across subtherapeutic, therapeutic, and supratherapeutic concentrations, indicating a sustained reduction in respiratory depression in favor of cebranopadol at all tested concentrations.

Fig. 5.

Comparison of the effect of cebranopadol versus oxycodone on ventilation at an extrapolated end-tidal carbon dioxide concentration (PCO₂) of 55 mmHg (${\dot{\mathrm{V}}}_{\mathrm{E}}55$) at equianalgesia. A, Absolute values (green line and shading, cebranopadol mean ± 90% CI; red line and shading, oxycodone mean ± 90% CI). The two drugs differ at the P < 0.05 level. B, Change in ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ between the two treatments. Data are percentage of baseline ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ ± 90% CI. Positive values indicate that cebranopadol is superior to oxycodone. Over the given concentration ranges, cebranopadol had an approximated 25% respiratory advantage over oxycodone (P < 0.05) at equianalgesia. Equianalgesia was determined by considering the C₅₀ values (the concentration causing 50% of maximum effect) for respiratory depression and pain relief.

Discussion

We conducted a randomized, partial crossover, controlled trial in healthy human volunteers to determine possible differences between the respiratory depressant effects of the novel dual NOP–µ-opioid receptor agonist cebranopadol and the pure MOP receptor agonist oxycodone. We conducted two analyses. In the LMM of ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ (primary endpoint), we observed a significant main effect on treatment (P < 0.001). The post hoc comparison between the different doses of cebranopadol and oxycodone did not reach the level of significance except for the comparison of cebranopadol 600 µg versus oxycodone 30 mg (P = 0.022). We relate the inability to unearth a post hoc separation for other cebranopadol oxycodone comparisons to the partial crossover design and consequently insufficient overlap of subjects among treatment groups, high variability in respiratory responses, and PK issues. We chose for the current design as we anticipated that few subjects would complete a study with six visits.

The secondary PK/PD analysis gives us an indication of the potency differences: As determined from the C₅₀, cebranopadol is 490 times more potent than oxycodone in producing analgesia, but just 180 times more potent in producing respiratory depression. Consequently, in order to achieve comparable levels of analgesia, substantially less cebranopadol is needed, and this is accompanied by a reduction of respiratory depression of approximately 25% when compared to oxycodone, as depicted in figure 5.

Another observation is the difference in onset/offset times between the two treatments with a much slower onset/offset for cebranopadol than oxycodone: t½k_e0 cebranopadol, approximately 60 min, versus 8 min for oxycodone. The slow onset of respiratory effect will, in clinical practice, allow the accumulation of arterial PCO₂ that will offset the full expression of respiratory depression.²⁸ In addition, the offset time of cebranopadol analgesic effect was much slower than that of oxycodone, indicative that analgesia will persist longer for cebranopadol than for oxycodone, and hence a single treatment per day may suffice.

Preclinical studies suggest that cebranopadol produces less respiratory depression than pure MOP receptor agonists and that this respiratory depression will have an apparent plateau at high doses.^8,15 Modeling from a previous clinical study suggests that this effect will carry over to humans.²¹ To confirm this and construct a robust model of respiratory effects, we administered multiple doses of cebranopadol and oxycodone. In this study, however, no apparent plateau in the respiratory response was observed for either cebranopadol or oxycodone. In the case of cebranopadol, the results were consistent with a plateau, but this could not be definitively concluded in our study. This might be related to the need for higher cebranopadol doses to determine the presence of such an effect. Our current safety data indicate that higher dosing will not compromise the safety of the subjects.

Another cause for not observing an apparent plateau may be related to the pharmacokinetics observed in the current study. We detected high variability in the plasma concentrations with limited separation in cebranopadol plasma concentrations between the three doses administered. Particularly in the case of the 1,000-µg dose, average exposure was less than expected based on previous studies. The reasons for this are unknown but could be attributed to high variability in cebranopadol pharmacokinetics, which is known from other studies,²¹ or could have been due to the oral formulation used in the study, which consisted of re-encapsulated drug that may have caused more variable uptake and consequently lower bioavailability than anticipated when considering the literature.

Our study had an unbalanced design with a partial crossover, i.e., all subjects received four of six possible doses (including placebo; fig. 1). This was done to increase subject compliance and reduce the drug load to opioid-naïve subjects. These study complexities (re-encapsulation, unbalanced design) may have prevented the detection of an apparent plateau in the cebranopadol respiratory data.

The current PK/PD model parameters agree with earlier findings from our laboratory after oral cebranopadol and oxycodone treatment. The combined data indicate the slow onset/offset for cebranopadol with effect-site equilibration half-lives ranging from 1 to 2 h for pain relief, respiratory depression, and miosis, with potency values that were similar for miosis and pain relief but greater respiratory potency (factor 1:5 to 6). Interestingly, oxycodone displays potency differences for the three endpoints with the greatest potency for respiratory depression, followed by miosis and analgesia (factors 1:2:17).

We used a single extrapolated isohypercapnic endpoint, i.e. ventilation at an extrapolated end-tidal PCO₂ of 55 mmHg, ${\dot{\mathrm{V}}}_{\mathrm{E}}55$, as the main endpoint in our study.^23–25 We earlier showed that this parameter is most sensitive in terms of respiratory potency and onset/offset times compared to resting parameters (baseline ventilation, baseline end-tidal PCO₂) and slope of the HCVR.^22,24 As stated earlier, the use of ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ as biomarker of respiratory depression in quantitative studies on the effect of opioids on ventilatory control allows standardized comparison among opioids in the assessment of their safety.

All treatments were well tolerated by the opioid-naïve subjects. Particularly the respiratory data (effect on ${\dot{\mathrm{V}}}_{\mathrm{E}}55$ and baseline ventilation as observed by apneas and oxygen desaturation) exemplify the advantage of cebranopadol over oxycodone. Furthermore, in clinical practice, just one dose of cebranopadol will be required per day, while two or more doses will be required for oxycodone given the differences in onset/offset times for analgesia (2 h vs. 4 min). Thus, the differences in respiratory adverse events in favor of cebranopadol seen in this trial based on a 24-h observation after a single dose may understate cebranopadol’s safety advantages because multiple doses of oxycodone would be required to treat pain during this same time period. We relate the difference in respiratory safety to the slow onset of respiratory effects, NOP agonism, or some combination of both.

Cebranopadol exhibits strong binding affinity for both the NOP and MOP receptors, with inhibitory constants of 0.7 and 0.9 nM, respectively, compared to weaker affinities for the kappa-opioid and delta-opioid receptors (inhibitory constants, 2.6 and 18 nM, respectively). The rationale for developing drugs that target both traditional opioid and NOP receptors is that earlier animal studies showed that activating the NOP receptor mitigates side effects linked to MOP receptor activation, notably respiratory depression. For instance, a study in nonhuman primates demonstrated that cebranopadol, when administered subcutaneously, provides pain relief without affecting ventilatory control as measured by baseline ventilatory parameters.⁸ These findings align with earlier rat studies, which also observed that while both fentanyl and cebranopadol provided effective pain relief, fentanyl had a much stronger effect on respiratory depression.¹⁵ In fact, selective antagonism of the NOP receptor enhanced cebranopadol-induced respiratory depression, a result that was reversed by naloxone, a µ-opioid receptor antagonist that showed MOP receptor dependency of the effect.²⁰ Based on these findings, researchers concluded that cebranopadol shows potent analgesic effects with a more favorable side effect profile compared to pure MOP receptor agonists.²⁹ These studies suggest that with dual NOP–µ-opioid receptor agonists could represent a promising new class of pain relief medications, particularly in terms of reducing the risk of respiratory depression. The next step is to determine how our current findings and those from previous studies translate into clinical pain management in humans.

Evidently, the healthy volunteers do not represent the complex pain population in clinical practice. Hence, our findings in healthy volunteers need translation to clinical pain management in patients with specific comorbidities or different pain states (e.g., inflammatory, neuropathic).³⁰ It is of interest to determine the potential differences between experimental pain models and clinical pain scenarios. Further, in this experimental study, we do not address chronic administration effects, such as tolerance, dependence, or long-term safety. While clinical data show advantage with respect to abuse potential over other opioids,¹⁸ further studies are needed to address these issues.

In conclusion, in the current study, we applied two separate models to analyze our data, a LMM based on administered doses (primary analysis) and a PK/PD model based on concentrations (secondary analysis). Both models showed that cebranopadol caused less respiratory depression than oxycodone (main effect in the LMM, and differences in C₅₀ in the PK/PD analyses) and potent analgesia (PK/PD analysis). In addition, even at doses that produce comparable levels of respiratory depression, cebranopadol appears potentially safer as evidenced by fewer instances of desaturation and other respiratory adverse events (safety analysis).

Further studies are needed to come to a definite conclusion regarding an apparent plateau in the respiratory effects of cebranopadol. Given earlier animal data and the results of this and an earlier human study, it is valid to conclude that cebranopadol may have a likelihood of producing a ceiling in respiratory depression in humans, similar to buprenorphine, which is a NOP agonist, albeit to a lesser extent than cebranopadol.^8,10,31 Further studies to address this issue are needed.

Research Support

This study was funded by Tris Pharma (Monmouth Junction, New Jersey).

Competing Interests

Dr. Dahan received consultancy fees from Zimhi Pharma Inc. (Carlsbad, California) and Enalare Therapeutics (Princeton, New Jersey) outside the submitted work. and received an award from the U.S. Food & Drug Administration (Silver Spring, Maryland). Drs. Hackworth, Grieco, and Lesnick work for Tris Pharma (Monmouth Junction, New Jersey). The other authors declare no competing interests.

Reproducible Science

Full protocol available at: s.c.jansen@lumc.nl. Raw data available at: s.c.jansen@lumc.nl.

Supplemental Digital Content

Supplemental Digital Content, https://links.lww.com/ALN/E330

Supplementary Material

aln-144-0913-s001.pdf ^{(2.3MB, pdf)}