Analgesic Efficacy of Peripheral κ-Opioid Receptor Agonist CR665 Compared to Oxycodone in a Multi-modal, Multi-tissue Experimental Human Pain Model: Selective Effect on Visceral Pain

Eighteen healthy, nonsmoking, opioid-naïve, white male volunteers (age 19–43 yr, median age 25 yr, weight 62.4–94.5 kg, median weight 80.0 kg, body mass index 20.5–27.4 kg/m², median body mass index 23.6 kg/m²) were recruited to participate in this single-center, single-dose, randomized, double-blind, placebo- and active-controlled, double-dummy, three-way, crossover study. The randomization ensured that six subjects had oxycodone, six had CR665, and six had placebo in each of the three study periods. This ensured a balanced design, and none of the persons involved in performing the study was involved in the randomization. All subjects were informed about the risks of the study, gave their written informed consent before participating, and were paid for participating. Volunteers entering the study were in good health and had no residual pain complaints from any previous illness. The study was conducted in accordance with the Declaration of Helsinki on biomedical research involving human subjects. The study protocol was approved by the Regional Committee on Biomedical Research Ethics, Aalborg, Denmark (registration no.VN-20060021) and by the Danish Medicine Agency, Copenhagen, Denmark (reference number 2612-3145).

Preadmission medical and concomitant medication history, physical examination including vital signs (orthostatic blood pressure and heart rate measurements taken after the subject had been semirecumbent for 5 min and then after the subject had been standing for 3 min, respiratory rate, and body temperature), height, weight and body mass index, 12-lead electrocardiogram, clinical laboratory tests (hematology, chemistry, and urinalysis) and urine drug screening tests were performed.

Subjects meeting the preadmission criteria were then scheduled for a visit during which the different pain tests were administered to ensure that each subject could tolerate the tests. Subjects underwent a urine drug screen and a test for alcohol at check-in before each treatment period. Immediately before administration of the first dose of study medication, the subject received a concealed randomization assignment.

The following inclusion criteria were applied: (1) semirecumbent blood pressure (after resting for 5 min) between the ranges of 90–139 mmHg systolic (inclusive) and 50–89 mmHg diastolic (inclusive); (2) partners must consent to use a medically acceptable method of contraception throughout the entire study period; (3) no known allergies to any of the compounds used in the study.

The following exclusion criteria were applied: (1) any evidence of clinically significant hepatic, reproductive, gastrointestinal, renal, hematologic, pulmonary, neurologic, respiratory, endocrine, or cardiovascular system abnormalities, psychiatric disorders, or acute infection; (2) any esophageal disease or disorders; (3) any abnormality on the screening electrocardiogram; (4) confirmed screening of QTc (heart rate-corrected QT interval [Q and T peaks of the electrocardiogram]) greater than 450 ms or a history of additional risk factors for torsades de pointes  (e.g. , heart failure, hypokalemia, family history of long QT syndrome) or the use of concomitant medications that prolong the QT/QTc interval (Q and T peaks of the electrocardiogram/heart rate-corrected QT interval); (5) resting heart rate at screening of less than 45 or greater than 85 beats per minute; (6) greater than 20-mmHg systolic or greater than 10-mmHg diastolic drop in blood pressure or greater than 30-beats per minute increase in heart rate within 3 min of standing or symptoms of lightheadedness or dizziness or fainting upon standing; (7) hemoglobin less than 12.5 g/dl (7.8 mmol/l) or donated blood or plasma or blood loss of more than 400 ml within 4 weeks before dosing; (8) use or planned use of medication during participation in the study.

Equal numbers of subjects were randomly assigned to one of three treatment sequences (ABC, BCA, or CAB), and all subjects received the following three treatments.

Treatment A: CR665 intravenous solution, 10 mg/ml, 1.1-ml vial, total dose of 0.36 mg/kg administered as an intravenous infusion over 1 h at a rate of 25 ml/h, and an oral placebo solution consisting of a colored, flavored beverage. This dose is selected on the basis of a safety phase I trial.

Treatment B: Oxycodone, 15 mg oral liquid solution mixed with the colored, flavored beverage used for the oral placebo solution in Treatment A and a 1-h intravenous infusion of vehicle placebo solution.

Treatment C: 1-h intravenous infusion of vehicle placebo solution and oral placebo solution as used in Treatments A and B.

Subjects were admitted to the unit on the morning of dosing, and the multi-modal esophageal tube was placed before dosing. Subjects remained in a semirecumbent position for the first 12 h after the beginning of dosing. After the 12-h period, subjects were allowed to sit up, stand, or walk with assistance only if their sustained standing heart rate was less than 100 beats per minute. The next morning, 24 h after dosing initiation, the subject was released from the clinic if no symptoms were observed. Each treatment was followed by a 1- to 3-week washout period.

To preclude possible physiologic effects from aquaretic activity, a known pharmacological effect of peripherally selective as well as centrally acting κ-opioid agonists, subjects were required to consume at least 1,500 ml of water from −24 to −3 h, and at least 500 ml from −3 to −2 h before dosing. After dose initiation, subjects were required to consume at least 250 ml of water at 1.5 h, and subjects received intravenous fluids (saline) for 1.5 to 4 h as needed to fully replace urine loss.

During each treatment period, the following safety measurements were obtained:

• (1) Vital signs (semirecumbent and standing blood pressure and heart rate, respiratory rate, and body temperature) were obtained at the screening visit and then during the treatment period, as follows: predose, 30 min, and 1, 1.5, 2, 3, 4, 8, 12, 16, and 24 h (termination) after dose initiation.

• (2) Blood and urine laboratory evaluations, fluid balance monitoring, and cardiac evaluation were obtained before and throughout the treatment period.

• (3) Adverse events were recorded throughout the study.

• (4) Before treatments and at the end of the study, the following assessments were performed: physical exam, including vital signs (semirecumbent and standing blood pressure, heart rate, respiratory rate, body temperature), height (screening only), weight and body mass index (screening only); 12-lead electrocardiogram; clinical laboratory tests (hematology, chemistry, urine analysis); urine drug screen and alcohol test.

The different pain assessment parameters from skin, muscle, and visceral stimulation were measured before treatment and at 30, 60, and 90 min after drug administration.

The cutaneous pinching pain tolerance threshold was determined by pinching a skin fold on the volar forearm at 20 cm distal from the elbow with an electronic pressure algometer (Somedic AB, Hörby, Sweden). The two probes each had a surface area of 0.28 cm². The pressure was continuously increased at a rate of 30 kPa/s until the threshold was reached.

The pressure pain detection and tolerance thresholds were determined by an electronic pressure algometer (Somedic). A probe with a surface area of 0.28 cm²was pressed onto the supinator muscle on the left forearm at 10 cm distal to the elbow. The pressure was increased at a rate of 30 kPa/s until the pain detection or the pain tolerance thresholds were reached.

The electronic cuff algometer (Aalborg University, Aalborg, Denmark14,15) consisted of a pneumatic tourniquet cuff, a computer-controlled air compressor, and an electronic 10-cm visual analogue scale (VAS). The 0 and 10 cm extremes on the VAS were defined, respectively, as no pain and as the worst pain imaginable.

The compressor (Condor MDR2; JUN-AIR International A/S, Nørresundby, Denmark) was connected to an electric-pneumatic converter (ITV2030; SMC Corp., Tokyo, Japan) and controlled by a computer through a data acquisition card (PCI 6024E; National Instruments, Austin, TX). The subjects could immediately terminate compression by means of a hand-held pressure release button connected to the data acquisition card. The pain intensity was recorded continuously on the VAS and sampled at 100-ms intervals. The computer continuously controlled the compression rate to ensure a linear increase in pressure. The pneumatic tourniquet cuff was wrapped tightly around the gastrocnemius muscle.

The cuff was automatically inflated (compression rate 0.50 kPa/s). The subject was instructed to rate the pain intensity continuously on the VAS from the first sensation of pain, and the pressure continued to increase until the subject pressed the pressure release button again when the maximum pain tolerance threshold was reached.

The multi-modal esophageal probe16was inserted through the mouth and passed into the stomach. The probe was gradually retracted to identify the location of the lower esophageal sphincter as a zone of high resting pressure that decreases with swallowing. The bag was placed 7 cm proximal to the sphincter, and the probe was taped to the subject's cheek. Subjects remained in a semirecumbent position with the head tilted back 30 degrees.

The probe was 0.5 cm in diameter with a polyurethane bag attached to the distal end for mechanical and thermal stimuli. The bag was 40 mm in length and could be inflated with fluid and then maintained at a constant volume through a pair of infusion channels. The infusion channels were connected to a precision infusion-withdrawal pump (type 111; Ole Dich Instrument Makers, Hvidovre, Denmark) that was used to fill the bag at a volume rate of 10 ml/min. The bag completely enclosed a side hole, which was used for measurement of pressure within the bag. A temperature probe (PR Electronics, Rønde, Denmark) was used to continuously monitor the fluid temperature in the bag.

For all visceral stimuli, a single 10-point electronic VAS was used to assess nonpainful (VAS < 5) as well as painful sensations (VAS ≥ 5) in response to the experimental visceral stimuli.17The subjects rated the intensities of the nonpainful visceral sensations from 1–4. A rating of 5 was defined as the pain threshold. Thus, with increasing stimulus intensity, the subjects rated the painful visceral sensations from 5–10, where 7 was rated as moderate pain.

For mechanical stimulation, the esophageal balloon was filled with 37°C water at a constant infusion rate of 10 ml/min until the subjects reached the pain threshold (VAS = 5) and moderate visceral pain intensity (VAS = 7) ratings. The total volumes (ml) required to reach these ratings were recorded.

The heat pain stimulus was generated by recirculating temperature-controlled water (55°C) in the esophageal probe,16and the volume of fluid in the bag was held constant. Before recirculation, the bag was filled with a volume of water corresponding to a mechanical VAS = 3 rating (prepain) to ensure reliable mucosal contact. The perfusate temperature was increased gradually until a constant temperature of 55°C was obtained. This temperature was maintained for 90 s or until the pain detection threshold (VAS = 5) was reached. As a measure of the total thermal energy delivered to the tissue, the stimulus intensity was calculated as the area under the curve (temperature [°C]× time [s]) from start to end of the stimulation.16 

To correct for individual differences in baseline pain recordings, the change in stimulus intensity relative to baseline was calculated for each measure for each subject. The results are expressed as mean ± SD unless otherwise indicated. For overall statistical assessment of baseline-corrected stimulus intensities associated with the sensation and pain thresholds under investigation, two-way analysis of variance was used with the factors of drug and time. Tukey test was used for post hoc  analysis. The difference in urine level was analyzed by one-way analysis of variance with drug as a factor. P < 0.05 was considered significant. The software package Sigma Stat 3.0 (Systat Software, Inc., Point Richmond, CA) was used.

All included 18 volunteers completed the study. Table 1summarizes all adverse events, which were all mild in severity with the exception of one episode of moderate increased heart rate in the CR655 treatment period. The dysphoria was characterized as mild, intermittent, and only possibly related to the study medication. No serious adverse events were reported. One subject used a concomitant medication (Zyrtec®[cetirizine hydrochloride]; UCB Nordic, Denmark) (for allergy in the placebo period) during the study.

The number of volunteers who urinated during the first 2 h after drug administration was 13 of 18 for CR655, 3 of 18 for oxycodone, and 4 of 18 for placebo. The mean volume of urine collected during the entire session was 4,187 ml for CR655, 3,463 ml for oxycodone, and 3,546 ml for placebo. No differences were found among groups (P < 0.148).

The average baseline pinching pain tolerance threshold was 900 ± 319 kPa. There was a significant difference in the effect of drugs (F = 26.3, P < 0.001) (fig. 1). Further post hoc  analysis showed that oxycodone increased pinching pain tolerance threshold in comparison with placebo and CR665 (P < 0.001). CR665 decreased pinching pain tolerance threshold compared to placebo (P = 0.007) and oxycodone (P < 0.001).

The average baseline pressure pain detection and pressure pain tolerance thresholds were 475 ± 182 kPa and 716 ± 330 kPa, respectively. No significant difference in the effect of drug was found (F = 0.154, P = 0.858) and neither oxycodone nor CR665 was different from placebo.

The average cuff pressure-pain tolerance thresholds were 48.68 ± 13.05 kPa. There was a significant difference in the effect of drugs (F = 12.3, P < 0.001). Further post hoc  analysis showed that compared to placebo and CR665, oxycodone significantly increased the cuff pressure-pain tolerance thresholds (P < 0.001) (fig. 2).

Repeated visceral stimuli are known also from previous studies10to cause a reduction in threshold over time. The averaged baseline pain detection threshold (VAS = 5) and a moderate pain threshold (VAS = 7) to distension were 17.7 ± 8.1 ml and 24.0 ± 11.2 ml, respectively. For the VAS = 5 threshold there was a significant difference in the effect of drugs (F = 6.18, P = 0.003). Further post hoc  analysis showed that oxycodone was better than placebo (P < 0.001). For the VAS = 7 threshold, there was a significant difference in the effect of drugs (F = 9.15, P < 0.001). Further post hoc  analysis showed that oxycodone was better than placebo (P < 0.001) and that CR665 was better than placebo (P = 0.005) (fig. 3).

The average baseline thermal energy (temperature °C × time [s]) was 205.8 ± 141.3°C/s. There was a significant difference in the effect of drugs (F = 7.96, P < 0.001). Further post hoc  analysis showed that oxycodone significantly increased the threshold compared to placebo and CR665 (P = 0.002 and P < 0.001, respectively). CR665 was not different from placebo (fig. 4).

The present human proof-of-concept experimental pain study showed significant analgesic effects of the peripherally acting κ-opioid agonist CR665 on visceral pain and a paradoxical hyperalgesic action on skin pinching pain.

Oxycodone showed pronounced effects on pain from somatic as well as visceral structures supporting previous studies.10,18The side effects after oxycodone were typical of μ-opioid agonists.19,20 

In the clinic, visceral pain is often difficult to alleviate with morphine and other μ-agonists, and it has been shown that κ-opioid agonists can be effective.21,22,23,24κ-Opioid agonists like fedotozine25and asimadoline have been tested for analgesic activity against pain from colonic distension in patients with irritable bowel syndrome and functional dyspepsia21,22,23with reported clinical effectiveness. However, asimadoline caused slight hyperalgesia in nonvisceral postoperative pain.26In the current study, CR665 caused hyperalgesia to skin pinching. The hyperalgesic effect of κ-agonists on cutaneous pain has not been reported previously in any preclinical or clinical studies on κ-opioid agonist, but it is known that central effects of dynorphin A (endogenous κ-opioid agonist) possess pronociceptive properties.27 

The side effects of CR665 were mainly limited to mild pruritus at the site of administration and to mild facial tingling (paraesthesia). Facial paraesthesia could be associated with κ-receptor activation.28 

One example of dysphoria was observed after CR665, which could be associated as a central effect; in general, CR665 showed less centrally related effects as observed with centrally acting κ-opioid agonists (e.g. , Pande et al.  29).

However, the dysphoria was characterized as mild, intermittent, and only possibly related to the study medication. However, other central nervous system side effects were reported (paraesthesia, somnolence, and dizziness), but it is not known whether the drug can cause central actions. Studies in humans with enadoline (CI-977),30a highly selective and potent κ-opioid agonist, have shown analgesic effects, although associated with neuropsychiatric side effects.29 

It is evident that peripherally located κ-receptors are of importance, particularly for visceral pain.3,31Clinical studies have been performed with a number of peripherally acting (and usually also, to some degree, centrally acting) experimental κ-opioids,32including asimadoline (EMD61753),33RP 60180,34niravoline (RU 51,599),35GR 94839,36ICI 204448,37enadoline (CI-977),38fedotozine,39spiradoline (U62,066E),40and ADL-10-0101.23 

These clinically tested κ-opioids have all produced unacceptable central side effects (e.g. , Pfeiffer et al.  41) at analgesic doses or unreliable efficacy that may be related to off-target activity (e.g. , Machelska et al.  26and Coruzzi et al.  42) and/or low affinity for the κ-opioid receptor (e.g. , Allescher et al.  43). These clinical reports, together with physiologic evidence for a role of κ-opioid receptors in modulating visceral pain,44,45have lead to an intensified search for high-affinity, peripherally selective κ-opioids.

Peptidic κ-opioids, e.g. , SK-9709,46E-2078,47FE200041 (an earlier analog of CR66548), and FE200665 (CR665, designated JNJ-3848850212,13) have been developed with the idea that a peptide should be less likely than a nonpeptidic small molecule to cross the blood-brain barrier and cause centrally mediated side effects. This strategy is not invariably effective: E-2078, a synthetic, all-l-amino acid analog of dynorphin A,1–8was found to readily cross the blood-brain barrier and produce apparently nonopioid analgesia in monkeys in doses overlapping with the sedative dose range.49However, preclinical studies with FE200041, an all-d-amino acid tetrapeptide κ-opioid agonist, demonstrated antinociception that was confirmed, with suitable antagonists to be peripherally mediated and selective via κ-opioid receptors.48Subsequent evaluation of an improved analog, FE200665 (CR665), confirmed this profile of peripheral κ-opioid selectivity and demonstrated peripheral antinociceptive activity in a wide range of preclinical visceral pain models.12,13The current experimental pain study with CR665 substantiates these animal data in man.

The need to improve the characterization of new compounds for the treatment of pain has led to development of a comprehensive battery of multi-modal, multi-tissue experimental pain models.8,9Collectively, this experimental concept has provided a valuable tool for differentiating visceral pain from other forms of pain and enables the profiling of new compounds such as CR665. This battery has recently been used to explore the differential analgesic effects of 15 mg of oxycodone and 30 mg of morphine10; these opioids were found to be equipotent in somatic pain (skin and muscle), whereas oxycodone was clearly more effective than morphine in visceral pain.10This study was then repeated in patients with chronic pancreatitis, and the efficacy of oxycodone on the various experimental pain parameters was found to be significantly enhanced18possibly due to the inflammatory components of the pancreatitis. On the basis of these observations, as well as substantial preclinical evidence that peripheral opioid analgesia is enhanced in the presence of inflammation, the rational next step in the development of CR665 is to apply it to patients with inflammatory visceral pain.

Although oxycodone has been generally considered to act as a typical μ-opioid agonist in humans, preclinical studies indicate that the antinociceptive effects of oxycodone are mediated by a combination of μ-opioid and κ-opioid receptors.11,50,51,52Consistent with this view, morphine-tolerant rats continue to exhibit analgesia with oxycodone, whereas oxycodone-tolerant rats fail to display analgesia with morphine.53Recently, it has also been shown that oxycodone and morphine have distinctly different pharmacological profiles in rat models of neuropathic pain.54Nevertheless, it is difficult to attribute these findings to κ-opioid receptor binding activity of oxycodone (receptor affinity greater than 1000 nm) or its metabolites (Staahl et al.  55). In human subjects, the only two metabolites with significant κ-opioid receptor binding activity are oxymorphone (receptor affinity 148 nm) and noroxymorphone (receptor affinity 87 nm), but these compounds are still 15-fold more potent at μ-opioid receptors.56In addition, of these two metabolites, noroxymorphone reaches the higher maximum plasma concentration (7.8 ng/ml at about 1.5 h), which is still about five-fold lower than the concentration of oxycodone itself, which is very potent at the μ-opioid receptor (receptor affinity 16 nm), but without activity at the κ-opioid receptor at 1,000 nm.56Clearly, then, the μ-opioid activity of oxycodone and its metabolites is collectively much greater than κ-opioid activity. However, this imbalance may not completely preclude some contribution of κ-opioid activity because interactions between μ-opioid and κ-opioid receptors appear to play a significant role in how nociception is mediated.57 

In patients with severe pain originating from the gastrointestinal tract, nonopioid analgesics are often insufficient to relieve pain to an acceptable level.58Treatment with traditional μ-opioid agonists also often fails to relieve the pain sufficiently, and at the same time causes constipation, which itself can be painful. To address these problems, new therapeutic approaches have emerged, and the development of new types of opioids that interact with different opioid receptors. CR665 and oxycodone differ from traditional μ-opioid agonists50and have shown significant effects on visceral pain and may hence provide new opportunities in the management of clinical visceral pain.