Effect of pretreatment with hydromorphone or buprenorphine on thermal antinociception induced by fentanyl in awake cats

Abstract

Objectives

The aim of this study was to determine the effect of pretreatment with hydromorphone or buprenorphine on thermal antinociception induced by fentanyl in cats.

Methods

Seven healthy cats received five different treatments consisting of two drugs. Drug 1 was administered intravenously 1 h before drug 2. Drug 2 was given as an intravenous loading dose followed by an infusion for 4 h. The drug combinations used were: buprenorphine 20 µg/kg followed by fentanyl (BF; 7 μg/kg, 7 μg/kg/h), buprenorphine 20 µg/kg followed by 0.9% saline solution (B), hydromorphone 0.07 mg/kg followed by 0.9% saline solution (H), hydromorphone 0.07 mg/kg followed by fentanyl (HF; 7 μg/kg, 7 μg/kg/h) and 0.9% saline solution followed by fentanyl (F; 7 μg/kg, 7 μg/kg/h). Thermal thresholds were obtained prior to treatment and at predetermined times up to 7 h after drug 1.

Results

Thermal thresholds were significantly higher than baseline in all treatment groups as follows: H from 0.25 to 2.50 h; B from 0.25 to 2.50 h; F from 1.25 to 5.50 h; HF from 0.25 to 5.50 h and BF from 0.25 to 5.25 h. Thermal thresholds were significantly higher in HF and BF than in F treatment before the fentanyl infusion was commenced (from 0.25 to 1.00 h). During the fentanyl infusion thermal thresholds in BF compared with F were lower at 1.75 h and from 2.50 to 3.50 h. After the constant rate infusion of fentanyl was started, thermal thresholds were significantly lower in HF compared with F at one time point (3 h).

Conclusions and relevance

Pretreatment with buprenorphine did partially inhibit the antinociceptive action of fentanyl. Hydromorphone did inhibit the antinociceptive action of fentanyl at one single time point in cats.

Introduction

Buprenorphine is considered a partial µ-opioid receptor agonist and the most popular opioid used in small animal practice in most European countries and Australia.^1,2 Buprenorphine is deemed to be preferable over other opioids for perioperative pain control in cats because it has a long duration of action and in contrast to hydromorphone and morphine, nausea, vomiting and salivation are uncommon after buprenorphine administration.^3–5 Buprenorphine has a very high affinity for the µ-opioid receptor. The drug’s unique receptor binding affinity suggests that full opioid agonists used to supplement analgesia with buprenorphine might be relatively ineffective. ⁶

Pure agonists, including hydromorphone and fentanyl, have high intrinsic activity at the µ-opioid receptor and therefore are effective analgesic drugs commonly used in dogs and cats for peri- and intraoperative pain management. ⁷

Several studies in different species have reported contradictory results about the interaction between buprenorphine and pure µ-opioid receptor agonists. Some results suggest that the combination of opioids with different µ-opioid receptor affinities may cause antagonistic effects. An in vitro study reported that buprenorphine administration suppressed fentanyl binding to µ-opioid receptors and suggested that the analgesic efficacy of fentanyl was reduced. ⁶ A clinical study in dogs advised to withhold buprenorphine therapy for 6–8 h before anesthesia incorporating pure µ-opioid receptor agonists as dogs pretreated with buprenorphine required more sufentanil injections intraoperatively. This suggests that the analgesic efficacy of sufentanil was reduced by buprenorphine. ⁸ Case reports in humans showed that pretreatment with buprenorphine hindered acute pain management with µ-opioid receptor agonists in trauma cases.^9,10

In contrast, in an experimental human pain model, the combination of fentanyl and buprenorphine did not lead to reduced analgesic action of fentanyl but led to an additive analgesic interaction. ¹¹ In mice, high doses of buprenorphine blocked antinociception of other opioids but low doses produced antinociceptive synergy. ¹²

The antinociceptive action of buprenorphine in cats has been intensively studied.^3,13–15 Currently, there are no investigations on the combination of buprenorphine with a pure µ-opioid receptor agonist administered to cats. The aim of this study was to investigate the effect of pretreatment with buprenorphine or hydromorphone on the antinociceptive efficacy of an intravenous (IV) infusion of fentanyl in awake cats. Furthermore, the time course of the analgesic effects and the behavioral side effects associated with drug administration were recorded.

The author hypothesized that administration of IV buprenorphine would inhibit the antinociceptive action of fentanyl, whereas pretreatment with hydromorphone would either have no effect or have an additive or synergistic effect on the antinociceptive effects of fentanyl.

Materials and methods

Animals

Seven healthy purpose-bred adult domestic shorthair cats (two intact females, five neutered males) were used in the study. Mean ± SD age was 1.6 ± 0.2 years. Mean ± SD body weight was 4.4 ± 0.8 kg. Based on physical examination, cats were deemed to be healthy. Results of a complete blood count and serum chemistry analysis were within accepted normal limits for the author’s laboratory.

All cats were housed in a climate-controlled room, in a group during the day and individually in a cage over night. Water was provided ad libitum and dry food was provided once daily. All cats were socialized and familiarized with the thermal threshold (TT) testing procedures and the testing environment for several weeks prior to study commencement. During the familiarization period, the TT equipment was positioned on the cats and the observer (BA) became familiar with the responses of individual cats to TT testing. This work was approved by the University of Saskatchewan’s Animal Research Ethics Board (AUP 20120104) and adhered to the Canadian Council on Animal Care guidelines for humane animal use.

Instrumentation and drug administration

On the morning of testing, cats were weighed and shaved over their lateral thorax for thermal threshold testing. A 22 G catheter (BD Insyte-W; Becton Dickinson) was placed aseptically in a cephalic vein. During the experiments all cats were allowed to move freely in a cage (115 cm × 75 cm × 85 cm) with unlimited access to food and water, a litter tray and toys, except that food was removed during infusion.

A blinded, randomized crossover study design with a minimum rest period of 7 days between treatments was used. A randomization chart determined the order of treatments. All cats received five treatments. The investigator (BA) who performed the experimental assessments was unaware of the treatment administered or the treatment order. Each of the treatments involved the administration of two drugs. Drug 1 was administered IV over 15 s 1 h before drug 2. Drug 2 was given as an IV loading dose (LD) over 5 s and immediately followed by a constant rate infusion (CRI) for 4 h. The drug combinations were: buprenorphine (20 µg/kg Vetergesic; Champion Alstoe Animal Health), followed by a LD of fentanyl (7µg/kg Fentanyl Citrate; Sandoz) and a CRI of fentanyl (7 µg/kg/h) (hereafter designated ‘BF’); saline solution (0.9% sodium chloride; Hospira), followed by a LD of fentanyl (7 µg/kg) and a CRI of fentanyl (7 µg/kg/h) (hereafter designated ‘F’); buprenorphine (20 µg/kg), followed by a LD of saline solution and a CRI of saline solution (hereafter designated ‘B’); hydromorphone (0.07 mg/kg Hydromorphone HCl; Sandoz), followed by a LD of saline solution and a CRI of saline solution (hereafter designated ‘H’); hydromorphone (0.07 mg/kg), followed by a LD of fentanyl (7 µg/kg) and a CRI of fentanyl (7 µg/kg/h) (hereafter designated ‘HF’). Drug 1 and the LD of drug 2 were diluted with 0.9% saline solution to reach a total volume of 0.15 ml/kg. All infusions were diluted with 0.9% saline solution and were administered by a computerized syringe pump (Baxter Model AS50 Infusion Pump; Baxter Healthcare) at the same volume and rate (2 ml/kg/h) administered at a set volume of 8 ml/kg over a period of 4 h.

Measurements of TT

Skin temperature and TT were determined four times at 15 min intervals prior to initiation of treatment and the mean value of each variable was calculated (baseline value). Drug 1 was administered at 0 h. Behavioral observations, sedation scores, and measurements of skin temperature and TTs were always obtained in the same order every 15 mins for the first 2 h and then every 30 mins for the remainder of the infusion period, then every 15 mins for first hour after the end of the infusion and then every 30 mins up to 7 h after administration of drug 1. For all experiments, TT was determined once at each time point by the same investigator (BA), who was unaware of the treatment administered. The degree of sedation was scored from 0 to 4 prior to TT testing, according to the following scoring system: 0 = cat displays normal behavior (ie, standing, walking, sitting, bright and alert, and interested in environment); 1 = cat is in a sternal or lateral position and stands when stimulated but is not interested in environment; 2 = cat remains in a sternal position, resists lateral recumbency but is not able to be aroused; 3 = cat remains in lateral recumbency but might lift head; 4 = cat remains in lateral recumbency even when stimulated, and is unable to rise or lift head. Behavioral and physiologic observations included signs of nausea (ie, salivation, lip licking), vomiting, mydriasis, and changes in physical activity and awareness.

For measurement of TT, a wireless testing system (WWT1; Topcat Metrology) was used. The TT system used in the study was previously developed and validated for the use in cats. It has been used in various investigations of the effects of analgesic drugs in this species. ¹⁶ The thermal stimulus was provided by a small probe that contained both a heating element and a temperature sensor. The probe was held in place against the shaved region of the cat’s thorax by an elastic hook and loop material band. A pressure bladder between the band and the probe was inflated to a known pressure to ensure even contact between probe and skin. The heater was activated via a handheld infrared remote control.

Before each testing, skin temperature was measured. When activated, the probe heated at a rate of 0.6°C/s with a safety cut-off value of 55°C to prevent skin damage. ¹⁶ During each measurement, the stimulus was terminated when the cat responded by jumping, flinching, turning towards the probe, licking or biting the probe area, or when the cut-off temperature was reached. The temperature at the termination of the stimulus was recorded as the TT at that time point. For testing, care was taken to ensure that cats were unrestrained and not sleeping, eating or playing.

Statistical analysis

A commercially available software package (GraphPad Prism 6.0; GraphPad Software) was used for statistical analyses. Normal distribution of each measured variable (skin temperature and TT) was confirmed using the Shapiro–Wilk test. Skin temperature and TT were compared between F and HF and F and BF by use of a two-way ANOVA followed by Tukey’s post-test for pairwise comparison. The within-group changes were examined with two-way ANOVA for repeated measures followed by a post-hoc Dunnett’s multiple comparison test to baseline. All data are reported as mean ± SD. P <0.05 was considered significant.

Results

Skin temperature did not vary significantly with time for either treatment. Overall skin temperature (mean ± SD) was 37.6 ± 0.3°C. No difference in baseline TT was detected between treatments. TT was significantly affected by treatment and time (Figures 1–4).

Figure 1

Thermal thresholds (mean ± SD) recorded in seven awake cats that received hydromorphone (0.07 mg/kg) followed by a constant rate infusion (CRI) of physiologic saline (H treatment; white squares) and saline (placebo) followed by a fentanyl CRI (7 µg/kg bolus, 7 µg/kg/h) (F treatment; black circles) in a randomized crossover design. Hydromorphone (H treatment) or placebo (F) was given 1 h before initiation of CRI treatments. Each CRI was maintained for 4 h.

Figure 2

Thermal thresholds (TT) (mean ± SD) recorded in seven awake cats that received hydromorphone (0.07 mg/kg) followed by fentanyl constant rate infusion (CRI; 7 µg/kg bolus, 7 µg/kg/h) (HF treatment; black/white squares) and saline (placebo) followed by a fentanyl CRI (7 µg/kg bolus, 7 µg/kg/h) (F treatment; black circles) in a randomized crossover design. Hydromorphone (HF treatment) or placebo (F) was given 1 h before initiation of CRI treatments. Each CRI was maintained for 4 h.

Figure 3

Thermal thresholds (mean ± SD) recorded in seven awake cats that received buprenorphine (20 µg/kg) followed by a constant rate infusion (CRI) of physiologic saline (B treatment; white circles) and saline (placebo) followed by a fentanyl CRI (7 µg/kg bolus, 7 µg/kg/h) (F treatment; black circles) in a randomized crossover design. Buprenorphine (B treatment) or placebo (F) was given 1 h before initiation of CRI treatments. Each CRI was maintained for 4 h. *Significant differences from baseline (Dunnet’s, P <0.05). ^†Significant differences between F and B (Tukey’s, P <0.05)

Figure 4

Thermal thresholds (mean ± SD) recorded in seven awake cats that received buprenorphine (20 µg/kg) followed by fentanyl constant rate infusion (CRI; 7 µg/kg bolus, 7 µg/kg/h) (BF treatment; white/black circles) and saline (placebo) followed by a fentanyl CRI (7 µg/kg bolus, 7 µg/kg/h) (F treatment; black circles) in a randomized crossover design. Buprenorphine (BF treatment) or placebo (F) was given 1 h before initiation of CRI treatments. Each CRI was maintained for 4 h. *Significant differences from baseline (Dunnet’s P <0.05). ^†Significant differences between F and BF (Tukey’s P <0.05)

In the H and B treatment groups, TT significantly increased from baseline from 0.25 to 2.50 h. In the F treatment group, significant increases in TT from baseline were recorded 15 mins after commencing fentanyl CRI and persisted for 30 mins after the end of the 4 h infusion (from 1.25 to 5.50 h). In the HF treatment group, significant increases in TT from baseline values were recorded 15 mins after hydromorphone administration and persisted until 30 mins after the end of the fentanyl CRI (from 0.25 to 5.50 h). The increases in TT from baseline values observed in the BF treatment group started 15 mins after buprenorphine injection and persisted throughout the duration of fentanyl CRI until 15 mins after the end of the infusion (from 0.25 to 5.25 h).

The TT was higher in the HF and BF treatment groups than in the F treatment group before the fentanyl CRI was commenced (from 0.25 to 1.00 h). After the CRI of fentanyl was initiated, TT values recorded in the HF and F groups were significantly lower in the HF group at one time point (3 h). During the fentanyl CRI in the BF treatment group the TT was significantly lower than the TT in the F treatment group at 1.75 h and from 2.50 to 3.50 h. The TT was higher in the H and B treatment groups than in the F treatment group before the fentanyl CRI was commenced (from 0.25 to 1.00 h). After the CRI of fentanyl was initiated, TT values recorded in the H and F groups were significantly lower in the H group at 1.75 h and from 2.50 to 5.00 h. During the fentanyl CRI in the B treatment group TT was significantly lower than the TT in the F treatment group at 1.25 and 1.75 h and from 3.00 to 5.50 h. The TT between H and B treatment groups did not differ.

Transient signs of nausea (lip licking) after hydromorphone administration were observed in three cats in the H group and in four cats in the HF group. In addition, one cat vomited and one salivated profusely initially after administration of hydromorphone in the H and HF treatment group, respectively.

In all cats marked mydriasis developed within minutes of the administration of hydromorphone, buprenorphine and fentanyl. Four cats in the H group and six cats in the HF group showed signs of sedation (sedation score 1) within 5 mins of hydromorphone administration. Duration of sedation was variable between cats but ranged from 15 to 30 mins. After the administration of buprenorphine five cats in group B, as well as in group BF, were sedated (sedation score 1) for 30 mins. One cat in group F was sedated (sedation score 1) for the entire infusion period of fentanyl. Three cats in group H, five cats in group B, five cats in group F, four cats in group HF and five cats in group BF had signs of euphoria, including increased rubbing and purring when interacting with humans and kneading with their forepaws. Signs of euphoria in groups H, B, HF and BF were observed following initial signs of sedation. After the LD administration of fentanyl two cats in group F, five cats in group HF and three cats in group BF became hyperactive for 15–45 mins. Hyperactivity included increased locomotor activity, cats became restless, increasingly playful and more affectionate when interacting with humans (increased rubbing, purring and rolling), but they did not resist handling.

The only adverse events related to the TT testing were small dermal lesions on the thermal testing area of two cats seen 24 h after testing. Skin lesions were observed in total three times and resolved in both cats prior to the next treatment.

Discussion

TT during the infusion period in cats receiving fentanyl were significantly higher at 1.75 h and from 2.50 to 3.50 h when compared with TT in cats receiving fentanyl pretreated with buprenorphine. This suggests that pretreatment with buprenorphine partially reduced the antinociceptive action of fentanyl in cats. Pretreatment with hydromorphone at the dose used decreased the antinociceptive effects of fentanyl at a single time point during the fentanyl CRI.

The results of this preliminary study should be viewed with respect to limitations in the study design. The first limitation is related to the small number of cats involved. It is possible that analysis of data from a larger group of cats would result in the detection of significant difference between treatments for a more extended duration. A posteriori power analysis revealed that approximately 12 cats used in this study would have been needed to detect the observed difference, with a power of 0.8 and significance level set at a value of P <0.05.

Another limitation of the present study is related to the drug doses used. Fentanyl, at the dose used in this study, produced near maximal effects on TT. This decreased the likelihood of detecting a significant positive antinociceptive interaction between fentanyl and hydromorphone. Fentanyl in the cats in this study was combined with a single, clinically relevant dose of buprenorphine or hydromorphone. The aim of the present study was to investigate the antinociceptive efficacy of opioid combinations and dosages commonly used in clinical practice. The results of other studies indicate that the antagonistic effects of buprenorphine on µ-opioid receptor agonists appear to be dose dependent.^14,15 In the present study the interaction may have been different at lower doses of buprenorphine and fentanyl, and different timing of drug administration. Thus time- and dose-related properties in this experiment might account for the lack of positive or negative interaction of fentanyl with hydromorphone or buprenorphine. The feline TT model was used in this study to test for antinociception. The TT system has been validated for studies of opioid antinociception in cats.^3–5,13,15 Pain, however, is a multifactorial entity and cannot be reproduced by a single experimental method. Previous studies have indicated that a mechanical threshold system is able to detect opioid-induced antinociception in cats.^17,18 A mechanical stimulus would stimulate different nociceptors than a thermal stimulus. ¹⁸ The incorporation of a mechanical threshold system in the present study might have produced more robust results. Clinical analgesia, however, might differ in intensity and duration from antinociception tested in experimental studies. Clinical studies are necessary to test if the antinociception produced in the present study correlates to clinical analgesia in cats.

In the present study, IV hydromorphone provided antinociception for 150 mins (2.50 h); these findings are in agreement with the results of another study in cats using the TT model where IV hydromorphone was used. ¹⁹ IV hydromorphone 0.05 mg/kg or 0.1 mg/kg demonstrated antinociception for 5–80 mins and 5–200 mins, respectively. ¹⁹

A significant increase in mean TT after buprenorphine administration was observed between 0.25 and 2.50 h in the cats in the present study. The antinociceptive effects of buprenorphine in cats have been extensively studied using the TT model developed and validated for cats. ¹⁶ Between these studies a large variability regarding duration and magnitude of the antinociceptive effects of buprenorphine exists. Steagall et al reported a significant increase in TT after buprenorphine (20 µg/kg, IV), from 15 to 480 mins, in six cats, ³ whereas Hedges et al reported only a transient increase in TT of limited magnitude (from 7 to 104 mins) after buprenorphine (20 µg/kg, IV) in six cats. ¹⁵ The reasons for the difference in antinociception after IV administration of buprenorphine in cats between previous studies and the present study are not entirely clear. The results of the present and aforementioned studies indicate that the effects of buprenorphine on TT are also highly variable between individual cats. This large variation in individual response resulted in a large SD when individual results were grouped by treatment. In the present study, this variation in TT between individual cats was also observed after treatment with hydromorphone and fentanyl but not for baseline thresholds. This suggests that variability in individual response to TT testing is likely to be associated with opioid treatment. Significant interindividual differences following administration of morphine have also been reported in a clinical study in humans. ²⁰ Individual response to opioid treatment is likely to be caused by genetic variability. Great individual variability with respect to morphology, number and distribution of opioid receptors has been described in dogs and horses. ²¹ Recent advances in molecular biology illustrate that a single µ-opioid receptor gene generates multiple µ-opioid receptor subtypes through splicing of mRNA. ²² The individual variation in µ-opioid receptor subtypes accounts for the variability in response to opioids because the actions of µ-opioids are the summation of their interactions with multiple µ-receptor subtypes. ²² Further research is necessary to confirm that cats possess similar pharmacogenetic variation, which would explain the interindividual variation in response to opioids observed in this and other studies.

In rats, co-administration of morphine and low-dose fentanyl produced long-lasting synergistic antinociceptive effects evaluated by the tail-flick test after single or repeated administration. ²³ Hydromorphone in our study did not increase the antinociceptive effects of fentanyl. There are several possible explanations for this lack of increased antinociceptive efficacy obtained with TT testing after the co-administration of hydromorphone and fentanyl in the cats in the present study. The absence of positive interaction in the present study is most likely related to the dose of hydromorphone and fentanyl used. It is known that synergy is not only a property of the drug pair, but also depends on the relative amounts in the combination tested. ²³ The challenge with opioid combinations is finding the right combination and the right dose for the right pain mechanism. ²⁴ The fentanyl infusion dose rate selected for the present study was similar to a dose rate commonly used at the author’s institution to provide analgesia during the perioperative period. This fentanyl dose rate produced near-maximal to maximal effects on TT. This reduced the probability for the pretreatment with hydromorphone to increase the antinociceptive effects of fentanyl and to detect a statistically significant difference between the two treatments. Further studies involving co-administration of hydromorphone with a low dose of fentanyl evaluated by a thermal and mechanical threshold testing are warranted to assess whether opioid combinations can be useful analgesics in cats. Another possible explanation might be that hydromorphone and fentanyl share the same mechanism of action through the activation of µ-opioid receptors at the same receptor sites after a thermal noxious stimulus. High selectivity but different affinity for the same receptor site may have resulted in competitive binding to the receptor site and subsequent lack of increased antinociceptive effects. This explanation seems unlikely in the light of the large number of different µ-opioid receptor subtypes. ²²

Pretreatment with buprenorphine reduced the antinociception of fentanyl at some time points in the present study. This suggests that buprenorphine at the dose studied 1 h prior to fentanyl administration partially inhibited the antinociceptive effects of fentanyl. Buprenorphine has high affinity for the µ-opioid receptor and low intrinsic activity in test tube assays.^25,26 Buprenorphine has been classified as a ‘partial agonist’ at the µ-opioid receptor because it does not elicit a maximal clinical response. ²⁷ The high affinity of buprenorphine for the µ-opioid receptor suggests that buprenorphine might reduce the antinociceptive effects of a full µ-opioid agonist when used concurrently. ⁶ The interaction between buprenorphine and pure µ-opioid receptor agonists has been studied in various species in preclinical and clinical settings but the findings of different studies contradict each other. Previously published work in dogs undergoing ovariectomy showed that buprenorphine reduced the antinociceptive effects of the µ-opioid agonist sufentanil. ⁸ Another clinical study in dogs undergoing thoracotomy suggested that the interaction between buprenorphine and the µ-opioid receptor agonist fentanyl were insignificant. ²⁸ Discrepancy between findings about the interaction of buprenorphine with µ-opioid agonists also exists in experimental studies.^11,12 Several studies have shown that combining buprenorphine with a full µ-opioid receptor agonist provided greater antinociception and analgesia when compared with either drug alone.^11,12 The combination of low-dose fentanyl and low-dose buprenorphine in an experimental human pain model led to an additive analgesic interaction. ¹¹ High doses of buprenorphine blocked antinociception of other opioids in mice but buprenorphine doses in the antinociceptive dose range resulted in an increased antinociceptive effect. ¹² This positive interaction was observed irrespective of whether buprenorphine was administered before or after the other opioid.

Conclusions

Findings of the present study indicate that at the doses studied hydromorphone decreased the antinociceptive efficacy of a fentanyl infusion at a single time point in awake cats. Hydromorphone appears to be a suitable pretreatment 1 h prior to fentanyl infusion in cats. Based on the results of this study, no clear recommendation can be made about the timing and dose of buprenorphine administered prior to a fentanyl infusion in cats. Results indicate that pretreatment with buprenorphine at the doses studied partially reduced the antinociceptive effects of fentanyl infusion in awake cats. Further research with a larger number of animals is required to investigate the antinociceptive interaction of different doses of buprenorphine with fentanyl infusions in cats.