An experimental study comparing the respiratory effects of tapentadol and oxycodone in healthy volunteers

Abstract

Background

There is a clinical need for potent opioids that produce little or no respiratory depression. In the current study we compared the respiratory effects of tapentadol, a mu-opioid receptor agonist and noradrenaline reuptake inhibitor, and oxycodone, a selective mu-opioid receptor agonist. We hypothesize that tapentadol 100 mg has a lesser effect on the control of breathing than oxycodone 20 mg.

Methods

Fifteen healthy volunteers were randomized to receive oral tapentadol (100 and 150 mg), oxycodone 20 mg or placebo immediate release tablets in a crossover double-blind randomized design. The main end-point of the study was the effect of treatment on the ventilatory response to hypercapnia and ventilation at an extrapolated end-tidal P_CO2 of 7.3 kPa (55 mmHg, VE55); VE55 was assessed prior and for 6-h following drug intake.

Results

All three treatments had typical opioid effects on the hypercapnic ventilatory response: a shift to the right coupled to a decrease of the response slope. Oxycodone 20 mg had a significantly larger respiratory depressant effect than tapentadol 100 mg (mean difference −5.0 L min⁻¹, 95% confidence interval: −7.1 to −2.9 L min⁻¹, P<0.01), but not larger than tapentadol 150 mg (oxycodone vs. tapentadol 150 mg: P>0.05).

Conclusions

In this exploratory study we observed that both tapentadol and oxycodone produce respiratory depression. Tapentadol 100 mg but not 150 mg had a modest respiratory advantage over oxycodone 20 mg. Further studies are needed to explore how these results translate to the clinical setting.

In contemporary medicine, opioids are the most important drugs to treat moderate to severe acute and chronic pain. However, opioids (mainly mu-opioid-receptor, (MOR), agonists) come with serious side effects, including abuse, addiction, nausea/vomiting, constipation, dizziness, sedation and respiratory depression.^1,2 Especially the combination of high abuse potential and the probability of respiratory depression is potentially lethal.³ Given these facts, there is an ongoing search for new potent opioids with less abuse potential and less respiratory depression than current agents.⁴ Tapentadol is a relatively new potent opioid for treatment of moderate to severe nociceptive, neuropathic and cancer pain.^5–7 In contrast to the classical MOR agonists, tapentadol has a dual mode of action, i.e. activation of the MOR and inhibition of the neuronal reuptake of noradrenaline.⁸ Importantly, the affinity of tapentadol for the MOR is relatively low, with a 50-fold lower affinity than morphine.⁸ Due to the lower opioid load, this dual mode of action has an important advantage. Analgesia is maintained despite the low MOR affinity due to the synergistic interaction between the two mechanisms. However, the affinity at the MOR coincides with fewer side effects than opioids with a higher MOR affinity, such as oxycodone. This has been observed for gastrointestinal side effects and tendency towards drug shopping (i.e. drug likability).^5,7,9 Whether this also holds true for the respiratory effects of tapentadol has not been studied yet.

The current exploratory study was designed to compare the effects of tapentadol 100 and 150 mg and oxycodone 20 mg on isohypercapnic ventilation (end-tidal P_CO2 concentration 7.3 kPa or 55 mmHg; VE55). We deduced from the literature that tapentadol 100 mg and oxycodone 20 mg are equianalgesic and hypothesized that tapentadol 100 mg is superior to oxycodone 20 mg in producing less respiratory depression as assessed by VE55. All treatments were oral, using the immediate release formulations of tapentadol and oxycodone.

Methods

Subjects

This single-centre 4-way crossover study was performed from December 2015 to November 2016 at the Anesthesia & Pain Research Unit of the Department of Anesthesiology at LUMC. The protocol was approved by the local Institutional Review Board (Commissie Medische Ethiek, Leiden, The Netherlands) and the Central Committee on Research Involving Human Subjects (CCMO) in The Hague. Participants were recruited by advertisement in the local newspaper and flyers posted within the facilities of Leiden University. All subjects gave written informed consent prior to enrolment in the study. After receiving informed consent, the subjects gave their medical history and a physical examination was performed. An independent physician performed the screening. The study was registered at trialregister.nl under identifier NTR5076.

Inclusion criteria were: age 18–45 years, body weight 50–100 kg and a body mass index between 18 and 35 kg m⁻², ability to read and understand the written informed consent form. Exclusion criteria included any medical, neurological or psychiatric illness, any known allergies to study medication, illicit drug use, a positive urine dip-stick drug test (testing for opioids, cocaine, amphetamines, cannabinoids, benzodiazepines) on the day of screening or morning of the study (Alere Toxicology Plc., Oxfordshire, UK), a positive alcohol breath test on the day of screening or the morning of the study, alcohol use exceeding 21 units per week or a history of alcohol abuse, a positive pregnancy test on the morning of the study, breast feeding, participation in an investigational drug trial in the three months prior to administration of the initial dose of study drug or participating in a trial more than 4 times per year.

Intervention

On one of four different occasions, subjects received a capsule containing tapentadol 100 mg (immediate release (IR) formulation; Grünenthal GmbH, Aachen, Germany), tapentadol 150 mg (IR formulation; Grünenthal GmbH, Aachen, Germany), oxycodone 20 mg (IR formulation; Mundipharma Pharmaceuticals BV, Hoevelaken, The Netherlands) or placebo (cellulose capsule fabricated by the local pharmacy). Each participant randomly received all 4 treatments on four separate occasions at least one week apart. The medication was ingested with 100 mL non-carbonated water. Treatment was given after all pre-drug control (baseline) data were collected.

The 100 mg tapentadol dose was based on previous studies on the efficacy of tapentadol in postsurgical dental pain.¹⁰ An oral dose of 100 mg IR produced effective pain relief for 4 to 6 h with a peak effect after 60–90 min. The dose of oxycodone is based on the 5-fold greater analgesic potency of oxycodone compared to tapentadol (Grünenthal GmbH, data on file).¹¹

Measurements

Respiration

Steps in end-tidal P_CO2 (P_ETCO₂) were applied using the “dynamic-end-tidal forcing” (DEF) technique. The DEF technique enables changes in P_ETCO₂ while maintaining the end-tidal oxygen concentration at a constant normoxic level. The technique is described elsewhere to extent.¹² In brief, subjects breathed through a facemask that was attached to a pneumotachograph/pressure transducer system (#4813, Hans Rudolph Inc., Kansas City, MO) and to three mass flow controllers (Bronkhorst High Tech, Veenendaal, The Netherlands) for the delivery of oxygen, carbon dioxide and nitrogen. The mass flow controllers were steered by a computer running the custom made RESREG/ACQ software. The software allows for the control of the end-tidal gas concentrations (by manipulation of the inspired gas concentration) and the simultaneous acquisition of respiratory data. The inspired and expired oxygen and carbon dioxide partial pressures were measured at the mouth using a capnograph (Datex Capnomac, Helsinki, Finland). Heart rate and arterial oxygen saturation (Sp_O2, Masimo pulse oximeter, Irvine, CA, USA) were measured continuously throughout the study day.

Respiratory measurements without any inspired CO₂ and hypercapnic ventilatory response (HCVR) curves were obtained at 7 separate periods: before any study medication was given and during the 6 h after intake of the medication at 1 h intervals. Measurements without added CO₂ included inspired minute ventilation, end-tidal P_CO2, end-tidal PO₂ and oxygen saturation. To obtain the HCVR, two to four steps in end-tidal P_CO2 were applied with step sizes of 0.6kPa (4.5 mmHg), 1.2 kPa (9 mmHg), 1.8 kPa (13.5 mmHg) and 2.4 kPa (18.0 mmHg) above resting P_ETCO₂ (Fig. 1). Each P_ETCO₂ step lasted 6–8 min, assuming at least 2-min of steady-state ventilation. The order of the steps was arbitrarily set. Throughout each HCVR, the end-tidal oxygen partial pressure was maintained at a normoxic level of 14 kPa (105 mmHg). Only in case of desaturations (Sp_O2<94%) supplemental inspired oxygen was given (FI_O2 0.5–1.0).

Nociception

An FPN 200 N Algometer (FDN 200, Wagner Instruments Inc., Greenwich, CT) was used to deliver pressure pain on an area of 1 cm² between thumb and index finger.¹³ The device has a force capacity (accuracy) of 200 (2) N or 20 (0.2) kgf and graduation of 2 N or 200 gf. A gradually increasing pressure were applied manually and subjects were asked to indicate when the procedure became painful (i.e. pressure pain threshold, PPT). At baseline all measurements were obtained in triplicate. Thereafter, measurements were obtained at 1-h intervals for the 6 h following drug intake.

Questionnaires

To get informed on the effect of treatment on sedation (alertness), dizziness (or lightheadedness), euphoria and dysphoria we queried the subjects regarding the presences of these symptoms by using a 11-point Likert scale ranging from 0 (=no effect) to 10 (=extreme effect) at 1-h intervals just before respiratory testing.

Adverse events

Adverse events (AEs) that were reported or observed were noted; these included nausea/vomiting, oxygen saturation drops below 94%, pruritus, agitation/anxiety and headache.

Randomization and blinding

Randomization was performed by the local pharmacy using a computer-generated randomization list. After subject allocation, the pharmacy prepared the medication. All drugs were re-encapsulated into small capsules that were identical in size, form, taste and colour. The tablets were packed in unmarked containers and delivered to the research team on the morning of the study day. The research team provided the participant with the unmarked tablet and checked the intake of medication. The research team remained blinded to treatment until the data analysis was complete (November 4, 2016). The study was independently monitored and data analysis was performed after the monitor had filed the final report ensuring that all Good Clinical Practice requirements were met.

Data analysis

Baseline respiratory data were collected on a breath-to-breath basis. The 2-min medians, obtained just prior to the HCVR, were used in the data analysis. The HCVR data (VI versus P_ETCO₂) were fitted to obtain the slope of the HCVR. The slope of the HCVR was estimated in R (The R Foundation for Statistical Computing, www.r-project.org, accessed 30 August 2017) using the median ventilation values of the last 2 min of each step in P_ETCO₂. From this analysis we calculated the main end-point of the study, VE55 or ventilation at 7.3 kPa or 55 mmHg, which was extrapolated from the HCVR, and takes both the slope and position of the HCVR into account and hence gives a better reflection of the respiratory effect of the intervention (Fig. 1).¹⁴

The study was initially powered on a 5.0 L min⁻¹ (SD 3.5 L/min) greater decrease in VE55 following oxycodone 20 mg intake relative to tapentadol 100 mg. At a power of 90% and an alpha of 0.05 this resulted in a sample size of 12. Since we had no a priori indication on either the magnitude of the difference in VE55 between these treatments or the respective variances we performed a per protocol interim analysis on VE55 after seven subjects completed the study. The interim analysis was done by an independent statistician and effect sizes were not reported to the study team. The effect size of (ΔVE55) was somewhat smaller than initially anticipated: 4 L min⁻¹ with SD 4 L min⁻¹. Fifteen subjects, rather than the 12 that were initially reported in the trial registry, were required to detect a significant difference of this magnitude at α = 0.05 and 1-β≥0.9 (see addendum in www.trialregister.nl; NTR5076, accessed 30 August 2017). The power and interim analyses were performed in SigmaPlot version 12.5 (Systat Software Inc., San Jose, CA). Subjects that did not complete the four experimental sessions were replaced by another subject. Their data were discarded.

Statistical analysis

A non-linear mixed effects analysis was performed on the main (VE55) and secondary endpoints (slope of HCVR, apneic threshold, resting ventilation, end-tidal P_CO2 and pain pressure threshold). Post hoc analysis was by χ²-test. To adjust for the multiple comparisons (interim and final analysis), P-values<0.01 were considered significant for the primary end-point (VE55); P<0.05 were considered significant for the secondary end-points. The analysis was performed in NONMEM version VII (ICON Development Solutions, North Wales, PA).

Longitudinal pharmacodynamic analysis

Additionally, the time-response data were analysed with a longitudinal pharmacodynamic model. In contrast to the conventional statistical analysis, the longitudinal takes into account the onset and offset times of the response as well as the effect size. The methodology of this analysis is given in the supplement to this paper.

Results

Twenty-one eligible subjects were approached to participate in the trial. Three subjects could not participate for logistic reasons. Eighteen subjects were randomized. Three subjects withdrew consent because of discomfort (n=2) or migraine during CO₂ inhalation (n=1). These three subjects were replaced by participants of the same gender that completed all 4 sessions. The data of the 15 subjects that completed the study were analysed. These 15 subjects had a mean age of 23 (range 21–31) years, mean weight 72 (59–90) kg, mean height 180 (162–197) cm and mean body mass index 23 (19–27) kg m⁻².

Examples of the effect of placebo, tapentadol 100 mg, tapentadol 150 mg and oxycodone 20 mg on the steady-state ventilatory response to carbon dioxide are given in Figure 1. The graphs show the individual data points (open circles) that make up the ventilatory response and regression curves (broken line through the data). Both panels show the two effects of opioids on the HCVR: a shift to the right coupled to a decrease of the slope of the HCVR. The intersection of the horizontal grey lines with the regression curves is ventilation at 7.3 kPa (55 mmHg) or VE55. It is clear that both tapentadol 100 mg and oxycodone cause a large reduction in VE55, while there is just a small effect on resting ventilation (the horizontal part of the response curves). The effect of all four treatments on VE55 and secondary end-points is given in Table 1.

Primary end-point

VE55. Both tapentadol and oxycodone produced significant and long-lasting respiratory depression as measured by VE55 following oral intake (P<0.01 vs. placebo; Fig. 2 and Table 1). For tapentadol, a dose-dependent effect was observed (the effect of 150 mg was 1.5 times larger than that of 100 mg tapentadol, P<0.01). Oxycodone 20 mg had a significantly larger respiratory depressant effect compared to tapentadol 100 mg (mean difference −5.0 L min⁻¹, 95% confidence interval: −7.1 to −2.9 L min⁻¹, P<0.01); the effects of oxycodone 20 mg and tapentadol 150 mg were of similar magnitude.

Secondary end-points

Respiratory depression from the two opioids was observed in resting ventilation, end-tidal P_CO2, slope of the HCVR and the apneic threshold, but not in oxygen saturation. Although more desaturations were observed in subjects on oxycodone treatment compared to tapentadol, see Table 2. For all secondary end-points, except Sp_O2, tapentadol dose-dependency was apparent; the effect of oxycodone 20 mg was similar to that of tapentadol 150 mg (Table 1).

The repeated pain measurements obtained at baseline allowed the estimation of the variability of the nociceptive response. The within-subject and between-subject variabilities were respectively 6% and 25% (% coefficient of variation). The average baseline pressure to evoke a pain response (pain threshold) was 43 (1) gf. The treatment-induced increase in pain threshold is given in Table 1. A significant increase greater than placebo was observed for all three opioid treatments with no difference between treatments. The peak increase in response above baseline was 6% (placebo), 15% (tapentadol 100 mg), 13% (tapentadol 150 mg) and 19% (oxycodone 20 mg).

Questionnaires

None of the treatments caused dysphoria; mild euphoria (scores<3) occurred in 3 subjects following oxycodone 20 mg and tapentadol 150 mg, while effects on dizziness and sedation were mild with maximum median effects no greater than 2 (dizziness) and 3 (sedation) on a scale from 0–10 (Fig. 3).

Adverse events

No unexpected adverse events did occur during the study. The most common adverse event was nausea that occurred in 12 subjects on 21 occasions. Additional adverse events that were noted were hallucinations (2 occasions in 2 subjects), itch (1 occasion) and the drop of Sp_O2 below 94% (6 occasions in 5 subjects). Desaturations were treated with supplemental oxygen. See Table 2 for the differentiation of adverse events per treatments. All adverse events resolved within hours.

Pharmacodynamic analysis

See the Supplementary materials for the results of the longitudinal pharmacodynamic analysis.

Discussion

This is the first study to directly compare the respiratory effects of oxycodone and tapentadol. In this experimental study we compared two doses of oral tapentadol (100 and 150 mg) and one dose of oral oxycodone (20 mg) in a group of healthy volunteers. Respiratory depression, as assessed by VE55, was observed following all three opioid treatments with quantitative differences between treatments. While greater respiratory depression was observed following oxycodone 20 mg compared to tapentadol 100 mg, oxycodone 20 mg and tapentadol 150 mg produced similar levels of respiratory effect.

Without exception all opioids that act at the mu-opioid receptor (MOR) produce depression of the respiratory centres in the brainstem.^1–3 Studies on opioid-induced respiratory depression (OIRD) show highly variable effects among the different opioids and comparisons between studies and opioids are hindered by differences in protocol, study population, study technique, doses, modes of administration, setting, et cetera.¹⁵ In the current study, we measured resting respiratory variables and additionally performed hypercapnic ventilatory testing. The hypercapnic ventilatory response is a highly sensitive tool for assessment of the respiratory effects of opioids.¹⁴ We used a steady-state isohypercapnic technique that produces data without interpretational difficulties (as compared to non-steady-state or rebreathing techniques).¹⁵ The main observation of this study is that all three opioid treatments affect the control of breathing (Table 1) with a reduction of the slope of the HCVR and a shift of the HCVR to the right. Parameter VE55 combines these two effects and is consequently a useful measure for the quantification of an opioid’s effect on the HCVR. Our data indicate a respiratory potency ratio of 7.5 between for oxycodone and tapentadol (similar levels of respiratory depression were observed at 150 mg of tapentadol and 20 mg of oxycodone). This ratio exceeds the commonly used ratio for analgesia of 5.0 (i.e. equinanalgesia to 20 mg oxycodone is obtained at 100 mg tapentadol) and hence suggests some advantage of tapentadol over oxycodone in this respect. However, although we observed equianalgesia between oxycodone 20 mg and tapentadol 100 mg in our study, we were unable to observe a tapentadol dose-related increase in analgesia. Whether this is related to the relatively small sample size of our study (the study was not powered to detect a difference in analgesic efficacy between treatments) or the limited sensitivity of the nociceptive assay (pressure pain) to tapentadol remains unknown. Assuming equianalgesia, our study does predict a respiratory benefit of tapentadol 100 mg over oxycodone 20 mg. Similar observations were made in the pharmacodynamic analysis (see Supplementary materials). To get an indication of the balance between respiratory depression and analgesia in our data set, we used the results of the pharmacodynamic analysis to construct utility or safety functions. The utility of drug effect is defined as the probability of obtaining the desired effect minus the probability of obtaining a side effect.^16–19 In Figure 4 we show the probability of at least 50% analgesia minus the probability of at least 50% respiratory depression and show a clear separation between the two tested opioids, with a predominant negative response for oxycodone (i.e. the probability of respiratory depression exceeds the probability of analgesia) and a zero response for tapentadol (i.e. similar probabilities for respiratory depression and analgesia). Since both doses of tapentadol were included in the pharmacodynamic model and consequently in the construction of the utility function, we cautiously contend that tapentadol produces less respiratory depression than oxycodone at equianalgesia.

Possible differences in respiratory effect may be related to the distinct mechanisms of action of oxycodone and tapentadol. Both opioids differ in their affinity for the opioid receptors. Oxycodone, available since 1917, is a relatively selective MOR agonist with a greater affinity for the MOR than the other opioid receptors.²⁰ Compared to oxycodone, tapentadol’s affinity for the MOR is about a factor of 5 lower while the affinity for the other receptors is of the same order of magnitude.⁸ The lower affinity of tapentadol for the MOR may be held responsible for its restricted side effect profile.

Apart from its effect at the MOR, tapentadol is a selective noradrenaline reuptake inhibitor. There is evidence from animal studies that increased extracellular concentrations of noradrenaline enhance descending inhibitory systems and potentiate pain relief.²¹ Since both adrenergic and opioid receptors are expressed within the respiratory network,²² it is of interest to discuss whether tapentadol’s adrenergic effect may have a stimulatory effect on breathing and counteracts some of the MOR-induced respiratory depression. Some indirect evidence comes from animal models of the RETT syndrome. RETT syndrome is a neurological disorder of genetic origin (mutations in the methyl-CpG binding protein 2 (MECP2) gene), which apart from mental retardation is associated with severe breathing irregularities.²³ A deficiency in noradrenergic modulation of the medullary respiratory network in Mecp2-/y mice (a mouse model of RETT syndrome) causes life-threatening breathing disturbance.²⁴ Furthermore, the HCVR is displaced to the right in this same mice strain due to a reduced chemosensitivity at low (1–3%) but not high (6–9%) CO₂ values.²⁵ Desipramine, a noradrenaline reuptake inhibitor, re-establishes low-P_CO2 chemosensitivity, restores the HCVR and improves breathing stability.^25–27 These data suggest that the excitatory effects of noradrenaline reuptake inhibition may offset part of the respiratory depression induced by opioid activation of mu-receptors in the respiratory network.

Our study has some limitations. (1) We did not measure plasma drug concentrations and consequently could not perform a proper pharmacokinetic-pharmacodynamic analysis. Although the longitudinal pharmacodynamic analysis provides information on the speed of onset and offset and potency of the respiratory responses (see Supplementary materials), it also hampered by the lack of pharmacokinetic data. A proper pharmacokinetic-pharmacodynamic analysis enables the quantification of opioid effect in terms of effect-site concentration. (2) Both tapentadol and oxycodone displayed a rather limited sensitivity in the single nociceptive assay (pressure pain) that we used in our study (Table 1). Animal data show that tapentadol has a high sensitivity in visceral, neuropathic and inflammatory pain models. Possibly different results are obtained when using more opioid-specific nociceptive tests. (3) This experimental respiratory study was performed under strict controlled conditions. Results from such studies do not always correlate with clinical observations. Respiratory observations in the clinical setting are therefore needed.

In conclusion, in this exploratory study we observed that both tapentadol and oxycodone produce respiratory depression. At equianalgesic doses a respiratory advantage of tapentadol over oxycodone was observed. Additional comparative studies are needed to further explore the respiratory effects of oxycodone and tapentadol. Such studies, alike the current one, are clinically relevant given the large number of patients that consume opioids. In the Netherlands alone almost 8% of the population use legally prescribed opioids for some form of chronic pain, with 50,000 new oxycodone prescriptions written every month.²⁸

Author’s contributions

RS (Rutger van der Schrier) wrote the protocol and performed part of the experiments;

KJ (Kelly Jonkman) performed the experiments;

MV (Monique van Velzen) supervised the experimental part of the study and helped in the writing of the manuscript;

EO (Erik Olofsen) performed the data analysis;

AMD (Asbjørn M Drewes) helped writing the protocol and manuscript;

MN (Marieke Niesters) supervised the project, helped in the writing of the protocol, participated in the experiments, data analysis, and helped writing the manuscript;

AD (Albert Dahan) wrote the protocol, performed data analysis, wrote the paper.

Supplementary material

Supplementary material is available at British Journal of Anaesthesia online.

Declaration of interest

A.D. and A.M.D. received consultancy and/or speaker fee from Grünenthal. A.D. is chairman of the Leiden University Medical Centre Medical Ethics Committee. He was not involved in the review of this study. None of the other authors report a conflict of interest.

Funding

This investigator initiated trial was supported in part by Grünenthal GmbH, Aachen, Germany.

References

Author notes