Comparison of Epidural and Systemic Tramadol for Analgesia Following Ovariohysterectomy

Introduction

In light of the necessity for pain control in animals, many strategies have been evaluated. Epidural opioids are among the best therapeutic tools because of their effectiveness and safety. Epidural opioids bind to presynaptic receptors on the spinal terminals of afferent neurons, inhibiting the release of excitatory neurotransmitters such as glutamate and substance P.¹ Epidural opioids also antagonize excitatory neurotransmitters by inhibiting the transmission of postsynaptic impulses in ascending pathways. Epidural opioids produce excellent and long-lasting analgesia and require smaller doses than systemic administration.¹

Morphine has been the most commonly used epidural opioid because its analgesia is longer-lasting than the more lipid-soluble opioids and its efficacy in the control of postoperative pain in dogs and cats has been reported by several authors.^2–4 In addition, morphine reduces minimum alveolar concentration (MAC) of inhaled anesthetics in dogs and cats.^5,6

Respiratory depression is epidural morphine’s most concerning undesirable effect in humans, which occurs when morphine is absorbed by blood vessels and is distributed to the central nervous system (causing early respiratory depression).⁷ Additionally, the rostral spread of the drug, which reaches the supraspinal opioid receptors in the areas of the brain that control respiration, can result in delayed respiratory depression.^7,8 This side effect is not frequently observed in dogs and cats; however, Troncy et al. (2002) reported mild respiratory depression in dogs after preemptive morphine epidural administration.⁴

Tramadol is a synthetic analog of codeine and has two complementary mechanisms of action that are distinct from those of pure μ opioid agonists. Tramadol has a weak affinity for the μ receptor, whereas the mono-O-desmethyltramadol (M1) metabolite has a higher affinity for the receptor.⁹ Additionally, tramadol causes the inhibition of norepinephrine and serotonin reuptake in the central nervous system, which may enhance the analgesic efficacy of tramadol.^10,11 McMillan (2008) confirmed that dogs are able to produce active tramadol metabolite, but the levels of M1 were low.¹²

There are few studies showing tramadol’s efficacy in dogs and cats, although one study reported an analgesic effect similar to morphine when tramadol was administered IV to bitches undergoing ovariohysterectomy.¹³ Tramadol significantly reduced the MAC of sevoflurane in dogs when administered as a continuous rate infusion.¹⁴ Several clinical studies using epidural tramadol in humans showed its analgesic efficacy and safety for the respiratory system and demonstrated better analgesia via the epidural route than IV.^15–19 Castro et al. (2009) showed that during the first 6 hr postadministration, epidural tramadol provided similar analgesia as epidural morphine to a standardized noxious stimulus in cats; however, epidural morphine outlasted the duration of tramadol.²⁰ Vettorato et al. (2010) investigated the pharmacokinetic profile of tramadol and its M1 metabolite in dogs after the either an IV or extradural injection of tramadol in dogs undergoing orthopedic surgery. That research group observed similar analgesia using the IV and extradural routes.²¹ They also showed that the pharmacokinetic profile was similar for tramadol and its M1 metabolite, irrespective of the route of administration.

In horses, both epidural tramadol and morphine induce long-lasting analgesia without excitation of the CNS.²² Epidural tramadol in dogs could certainly contribute to controlling postoperative pain with no risk of respiratory depression. Another advantage of tramadol is the absence of strict regulatory measures with regard to its use and the absence of preservatives.

The purpose of this study was to compare the effectiveness of epidural and systemic tramadol in the control of postoperative pain in bitches undergoing ovariohysterectomy. The hypothesis was that epidural tramadol would promote better and long-lasting analgesia than the systemic route.

Materials and Methods

Anesthesia and Surgical Procedures

All animals underwent elective ovariohysterectomy and were randomly assigned to one of two groups: the epidural (EPI) or intramuscular (IM) group. All dogs were premedicated with IM acepromazine^b (0.1 mg/kg). The cephalic vein was then canulated, and lactated Ringer’s solution was administered throughout the entire procedure at an infusion rate of 10 mL/kg/hr.

Twenty minutes after premedication, animals in the EPI group (n=10) were prepared for tramadol administration. The injection site was clipped and cleaned in an aseptic manner. All animals were administered a sedative dose of propofol^c (2 mg/kg IV) to facilitate epidural catheter placement. Supplemental oxygen was administered by mask. The epidural catheter was placed before general anesthesia to evaluate the anal sphincter’s relaxation and to confirm the proper position of the catheter. A small dose of lidocaine was administered epidurally to both groups (as described below) to ensure the drugs had been administered epidurally, which was confirmed by anal sphincter relaxation. The lumbosacral space was punctured with a Tuohy (19 SWG) needle^d with the animal positioned in sternal recumbency.²³ Proper needle positioning was confirmed by lack of resistance to injection of 1.0 mL of 0.9% NaCl and subsequent introduction of a 20 SWG epidural catheter^e threaded up to the fifth to sixth lumbar intervertebral space. The catheter position was radiographically verified at the end of surgery by injecting 0.8 mL of iohexol^f. All epidural catheters were placed by the same individual (S.M.) who evaluated animal pain. After placement, the catheters were protected with a sterile bandage^g that had an antimicrobial filter in its edge.

Animals in the EPI group received epidural tramadol^h (2 mg/kg) combined with 2% lidocaine 2% (1.25 mg/kg) diluted in 0.9% saline in a volume of 0.26 mL/kg. Each dog in the EPI group received 3 mL of 0.9% saline IM. Animals in the IM group received 2 mg/kg tramadol^h diluted to 3 mL IM and epidural lidocaine (1.25 mg/kg) diluted to a volume of 0.26 mL/kg. All dogs were maintained in sternal position for 30 minutes after administration of the epidural drugs to allow adequate spreading within the epidural canal. Animals were manually restrained so additional doses of propofol were not necessary.

Thirty minutes after epidural administration of the tramadol and/or lidocaine, induction of anesthesia was performed with propofol^c to effect (4–6 mg/kg), which was slowly administered IV in a single bolus. Orotracheal intubation was performed, and anesthesia was maintained with isofluraneⁱ in 100% oxygen via a small animal circle rebreathing circuit (the fresh gas flow rate was 1 L/min). All dogs were allowed to breathe spontaneously. Anesthesia was initiated with an end-tidal isoflurane concentration of 1.4% and adjusted (an increase or decrease of 10%) according to changes in blood pressure and/or heart rate (based on an increase or decrease of 20%) as well as conventional signs of anesthesia depth (i.e., loss of palpebral reflex assessed by gentle brushing the eyelashes of one eyelid with a finger and eye globe position). Animals were maintained at the surgical plane of anesthesia (stage III, plane 2).

During the anesthetic procedure, the end-tidal partial pressure of CO₂, inspired CO₂, end-tidal isoflurane, and peripheral arterial oxygen saturation of hemoglobin were continuously measured with a capnograph, gas analyzer, and pulse oximeter^j, respectively. Samples to measure end-tidal isoflurane were drawn from the proximal end of the endotracheal tube at a rate of 150 mL/min. The gas analyzer was calibrated at the start of each experiment using the calibration gases supplied by the manufacturer. These above-described variables were recorded at 40 min, 50 min, and 60 min after tramadol administration (Table 1). During the procedure, arterial pressure was measured invasively by a catheter (a 20 spring-wire guide for larger animals and a 22 spring-wire guide for smaller animals) placed (after orotracheal intubation) in the metatarsal artery. The catheter was connected to the pressure transducer (zero-calibrated at the level of the right atrium) of the multiparameter analyzer^a. The pulse and respiratory rates were measured continuously using the same monitor.

TABLE 1 Description of Time Points for Patient Evaluations

View Fullscreen

Blood samples (2 mL) were collected from the femoral artery in heparinized syringes for the measurement of pH and blood gases at -30 min, 50 min, 120 min, 240 min, and 1,440 min after tramadol administration. All samples were analyzed immediately after sample collection using a blood gas analyzer^k. Blood gases were corrected for body temperature. Duration of surgery, time to extubation (i.e., time from when isoflurane was turned off until the dogs swallowed), and quality of recovery (based on either the presence or absence of signs of excitation, including paddling, vocalization, and head movements) were all recorded subjectively. All animals were placed on a thermal mattress^l during surgery, and all surgeries were performed in the morning by the same surgeon (A.T.), who was also a senior investigator.

Postoperative Assessments

Analgesia, sedation, heart and respiratory rates, and arterial blood pressure were evaluated at 120 min, 180 min, 240 min, 300 min, 360 min, 720 min, and 1,440 min after tramadol administration. Two scales were used to measure analgesia. Analgesia was always measured by a senior observer (S.M.) who had already used a combination of scales to evaluate pain in previous studies.^13,25 The individual that performed postoperative analgesia evaluation was always the same and was blinded to the treatment groups. The scales were a numeric rating scale (NRS) described by Hansen (1997) and Hopkins et al. (1998), with a score of 0 meaning no pain and 10 being the worst possible pain, and the scale proposed by Firth and Haldane (1999) (Table 2).^26–28 A visual analog scale was used to record sedation, according to Brodbelt et al. (1997).²⁹ Evaluation of animals was performed 720 min after tramadol administration by owners who received the Melbourne scale. The owners were trained how to perform the evaluation and how to administer intramuscular tramadol at home in case of a high Melbourne score.

TABLE 2 Description of the Melbourne Scale (Described by Firth and Haldane, 1999) Used to Assess Pain Postprocedurally

*

The term guarding refers to the animal turning its head toward the affected area; biting, licking, scratching at the wound; snapping at handler; or tensing muscles and adopting a protective (guarding) posture.

†

Does not include alert barking

View Fullscreen

Animals returned to hospital 24 hr after the analgesic administration and were reevaluated by the same observer that performed evaluation during postoperative period in hospital.

Animals with an NRS >4 or a Melbourne score >12 were administered 2 mg/kg tramadol IM. Ketoprofen (2 mg/kg per os q 24 hr for 3 days) was prescribed to all animals at the end of the experiment (24 hr after initial epidural or IM tramadol administration). Any incidence in nausea or vomiting was also recorded.

Results

No significant differences were detected between the two treatment groups regarding age, weight, surgical time, and time to extubation after anesthesia (Table 3).There were no differences between the two treatment groups in either respiratory rate or systolic, mean, and diastolic arterial blood pressures measured during the evaluation period (Table 4). The pulse rate, however, showed a treatment effect in the early postoperative period (120 min) because the IM group had a higher value than the EPI group. Pulse and respiratory rates showed a significant decrease at 40 min, 50 min, and 60 min (P=0.0001) compared with data obtained prior to premedication in both groups. The same pattern was observed with mean blood pressure, which decreased significantly at 40 min compared with data obtained prior to premedication (−30 min). Arterial partial pressure of O₂ and peripheral arterial O₂ saturation of hemoglobin did not vary between groups or evaluation time points (Table 4). End-tidal isoflurane was not affected by treatment (P<0.005), and it was not noticed end-tidal reduction during the time points in both treatments; End-tidal isoflurane in the IM and EPI groups at 40 min, 50 min, and 60 min were 1.21 ± 0.42 and 1.25 ± 0.26, 1.42 ± 0.48 and 1.37 ± 0.38, and 1.20 ± 0.38 and 1.1 ± 0.19, respectively. There was no difference in pH between the two treatments; however, a significant increase in pH in both groups was noticed at 120 min compared the pH values measured 50 min after tramadol administration. A significant increase in the arterial partial pressure of CO₂ was recorded 50 min after tramadol administration compared with the −30 min value (Table 4).

TABLE 3 Comparison of Weight, Surgical Time, and Time to Extubation in the Two Treatment Groups

Data are presented as mean ± standard deviation.

View Fullscreen

TABLE 4 Summary of the Various Postoperative Parameters Prior to Preanesthetic Medication and in the Intraoperative and Postoperative Periods in Dogs Treated with Either Epidural Tramadol (EPI) or Intramuscular Tramadol (IM)

*

Significantly different from time −30 min

†

Significantly different from EPI

§

Significantly different from time 50 min

Data are presented as mean ± standard deviation. DAP, diastolic arterial pressure; MAP, mean arterial pressure; NA, not available/measured; PaCO₂, arterial partial pressure of CO₂; PaO₂, arterial partial pressure of O₂; PR, pulse rate; RR, respiratory rate; SAP, systolic arterial pressure; SatO₂, arterial oxygen saturation; Temp, temperature.

View Fullscreen

For both groups, quality of recovery was considered good. Signs of discomfort, excitement, nausea, and vomiting were absent.

During the evaluation period, both groups had similar pain scores (Figure 1, Table 5), and almost all dogs had low pain scores according to both scales. Two dogs in the IM group required supplemental analgesia at 120 min and 180 min after tramadol administration, and one dog in the EPI group required additional analgesia at 120 min. There were no differences in sedation scores during the postoperative period between the two groups. Significantly higher sedation scores were observed at 120 min in both groups (3.5 ± 0.8 and 2.6 ± 1.2 in the IM and EPI groups, respectively) compared with all other time points. Higher sedation scores were observed at 180 min (2.2. ± 0.13 and 2.0 ±1.1 in the IM and EPI groups, respectively) compared with 24 hr scores (when the score was 0 for both groups).

View Figure

• Download as Powerpoint
• Download Figure

TABLE 5 Postoperative Pain Scores Employing the Melbourne Scale of Dogs in the Different Groups After Ovariohysterectomy

Data are presented as mean standard ± deviation.

View Fullscreen

There was a significant increase in serum cortisol levels at 120 min and 240 min after tramadol administration in both groups compared with premedication values (Figure 2). After 24 hr of observation, both groups had serum cortisol values similar to those measured in the preoperative period. There were no differences between the two groups in plasma norepinephrine concentrations. In contrast, a significant increase of norepinephrine was detected (P<0.001) in both groups in the early postoperative period (120 min) comparison with the intraoperative period (Figure 3A). At 50 min, 120 min, and 1,440 min, significantly higher epinephrine concentrations were detected in the IM group than the EPI group (Figure 3B).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Discussion

The choice of tramadol as the target of this research is based on the satisfactory results observed in previous studies when the drug was systemically administered to dogs and epidurally administered to humans, cats, and horses.^{13,14,16,19,20,22} The use of epidural tramadol in dogs compared with the IM route in dogs undergoing abdominal surgery has not been reported before. No control group was used in the current study, which might have facilitated comparisons and interpretation of results. However, because ovariohysterectomy can cause a moderate degree of pain and no other agent with analgesic property was administered to the dogs included in this study, the authors elected not to include a control group. The dose of epidural tramadol administration was chosen based on the dose used in children.^19,30

Epidural analgesia with opioids like morphine and fentanyl (for example) is a very useful and effective strategy to relieve pain in dogs because epidural opioids possess advantages over systemic administration, including long-lasting analgesia, great potency, and few side effects.² In this study, both treatments showed satisfactory and similar analgesia, which was demonstrated by the absence of differences between pain scores values and the low necessity of rescue medication in both groups during the evaluation period. These results are in agreement with Vettorato et al. (2010) who observed that the postoperative analgesia obtained in dogs administered extradural tramadol was similar to dogs that were administered IV tramadol following orthopedic surgery.²¹ Further, Güneş et al. (2004) reported better and longer-lasting postoperative analgesia with epidural tramadol than IV tramadol in humans.¹⁹ One possible explanation for this discrepancy may be the type of procedure involved. Güneş et al. (2004) used tramadol for pain relief in boys undergoing hypospadias repair, whereas the current study involved dogs undergoing ovariohysterectomy. Natalini et al. (2000) observed satisfactory results using epidural tramadol in horses, as did Castro et al. (2009) in cats.^20,22 However, in those studies, epidural administration was not compared with the systemic route.

In the current study, lidocaine was unlikely to have influenced the pain results because the dose used was very low. Further, both groups received the same dose of lidocaine epidurally.

When tramadol is administered systemically, it is metabolized into mono-O-desmethyltramadol (the M1 metabolite), which has a much higher affinity for the μ receptor than the parent compound.^10,11 It seems that humans have higher M1 production compared with dogs. In fact, some studies demonstrated low levels of M1 in dogs.¹²

In this study, two pain scales were included. One scale was subjective (NRS) and the other was objective (Melbourne scale). Many other studies also described the use of more than one scale in similar research studies in dogs.^25,31,32 Although the NRS is widely used for pain evaluation, it is subjective and is therefore not considered the gold standard for pain evaluation. The Melbourne scale is more objective because it uses some physiologic data, such as heart and respiratory rates. Hellyer et al. (2007) claimed that the Melbourne scale is promising for clinical use.³³ The results obtained with the two scales were similar, showing that they are both useful tools for evaluating pain in dogs. Due to the subjectivity of the NRS, especially for research purposes, the use of two scales is recommended.

To evaluate pain, subjective parameters were recorded in addition to measuring blood cortisol and catecholamine levels. Serum cortisol is known as a reliable method to evaluate pain and stress in small animals.³⁴ For example, Church et al. (1994) showed that cortisol concentrations increased in dogs undergoing major abdominal, thoracic, or orthopedic surgery.³⁴ In the current study, a nonsignificant increase of serum cortisol was observed in both groups during anesthesia. Because isoflurane was being administered in a relatively low concentration (1× MAC), it is possible to infer that the opioid contributed to to the control of cortisol increment in both groups. According to the results observed here, none of the treatments modulated the cortisol response in the early postoperative period, a finding that differs from those in the study by Block et al. (2003).⁸ The findings by Block et al. (2003) showed better analgesia with the use of epidural opioids than the systemic route. Buback et al. (1996) and Mastrocinque and Fantoni (2003), however, observed satisfactory serum cortisol modulation after surgery by using systemic opioids in dogs.^13,35 In the investigation of Mastrocinque and Fantoni (2003), there was an increase in cortisol during surgery, with values of cortisol reaching 4 μg/dL.¹³ In the present investigation, the increase in cortisol was similar to Mastrocinque and Fantoni results, but was only observed during the early postoperative period. It is difficult to affirm that the route of tramadol administration influenced the timing of the cortisol increase.

Animals’ behavior and their responses to the environment can change their epinephrine and norepinephrine concentrations.³⁶ A significant increase in epinephrine values was observed only in the IM group. Benson et al. (1991) demonstrated that anesthesia did not affect epinephrine and norepinephrine levels in cats.³⁷ The same was observed in relation to norepinephrine in the current study. In a previous study that compared systemic opioids in dogs, catecholamine concentrations were not influenced by anesthesia.¹³ Another important point to consider is that the epinephrine production changes according to the type of surgery. During abdominal surgeries, for instance, manipulation of the adrenal gland may occur, leading to a greater change in epinephrine than norepinephrine concentrations, which could explain the more pronounced changes in epinephrine concentrations during surgery found in the current study.³⁸ Lin et al. (1993) observed catecholamine modulation in cats receiving systemic opioids for 4 hr during the postoperative period. In Lin et al. (1993) study, modulation of epinephrine was better with epidural tramadol than with systemic administration, although in relation to baseline values, no significant change was observed.³⁶ It is possible that epidural administration assured the swift and direct binding of tramadol to the opioid receptors in the spinal dorsal horn, which promoted activation of the inhibitory antinociceptive system and better modulated epinephrine release compared to IM administration.³⁹

Tramadol is described as an agent that is free of respiratory depression. In the present investigation, respiratory depression was minimal, based on either clinical normal values of respiratory parameters in dogs or values obtained in other studies that used morphine in similar methods.¹³ In fact, none of the treatments promoted significant changes in respiratory rate and the arterial partial pressure of CO₂ during the postoperative period; however, an increase in the arterial pressure of CO₂ during the intraoperative period compared with before premedication was verified. As observed in previous research in humans with epidural tramadol, the current study found that cardiovascular changes were also minimal in both groups, which implies that epidural and IM tramadol cause similar cardiorespiratory changes.^16,17 The only exception was pulse rate. At 120 min, the pulse rate was higher in the IM group. The pulse rate increase in the IM group was accompanied by an increase in cortisol (compared to baseline), norepinephrine (compared to 50 min) and epinephrine (compared to the EPI group at the same time points), which can indicate a stress response related to pain. At 120 min, two dogs required rescue analgesia, which is in accordance with the neuroendocrine response.

The reduction in the amount of inhalant anesthetic agent required during surgery contributes to cardiorespiratory stability.⁶ There are published data showing the decrease in inhalant agent requirement concentration in dogs receiving continuous infusion of tramadol.¹⁴ There were no differences between the two groups in relation to end-tidal isoflurane in this study; however, it was observed that both treatments showed isoflurane concentrations comparable to 1× MAC.⁴⁰ This result is in accordance with the result obtained by Ozcengiz et al. (2001) who recorded a decrease in end-tidal isoflurane when they used tramadol or morphine in children.³⁰

Propofol was used in a sedative dose to promote a light plane of anesthesia (as described by Almeida et al. [2007]) to minimize the stress response due to the anxiety and pain associated with epidural catheter placement.²⁵ Because propofol has a short period of action, a second dose was administered immediately prior to intubation.

Vercauteren et al. (1999) did not observe effective analgesia when they used epidural tramadol after performing cesarean sections in women, and a high incidence of nausea and vomiting was also recorded.⁴¹ These side effects were neither observed in any dog during the current study nor were they reported in a previous study using systemic tramadol in dogs following a single administration of tramadol.¹³

Conclusion

Both epidural and systemic tramadol were similarly effective and safe for treating postoperative pain in bitches undergoing ovariohysterectomy. The neuroendocrine response to pain was better modulated (but not completely abolished) by epidural tramadol than IM tramadol. In both groups, end-tidal isoflurane concentrations were similar, and anesthesia was accomplished with a value comparable to 1× MAC. Epidural tramadol did not show advantages with regards to analgesia and cardiorespiratory stability compared with IM administration.

The authors would like to thank the University Hormonal Analysis Laboratory of University of São Paulo for cortisol measurements and the University Clinical Laboratory of São Paulo of the University of São Paulo for the catecholamine measurement. The author would like to thank FAPESP for the financial support.