Severe pruritus and myoclonus following intrathecal morphine administration in a dog

Abstract

During epidural needle placement in a 32-kg dog the subarachnoid space was punctured and half the intended dose of lidocaine, bupivacaine, and morphine was injected. After recovery from anesthesia the dog showed signs of severe pruritus of the tail base and limbs and myoclonus of the tail and hind limbs. Methadone, acepromazine, ketamine, buprenorphine, and butorphanol were administered to control myoclonus and pruritus, but were unsuccessful. Diazepam was used to control myoclonus until the effects of morphine abated.

Intrathecal injection is not commonly performed as a local anesthetic technique in dogs, whereas epidural anesthesia is a widely used and an effective method of providing perioperative and postoperative analgesia. Inadvertent subarachnoid puncture occurs in 2% to 4% of cases during epidural anesthesia in small animals (1,2) and, if this occurs, a 50% reduction of the intended epidural dose is recommended (3,4).

Epidural local anesthetics provide profound intraoperative analgesia and are often combined with longer lasting agents such as morphine (4). Epidural morphine provides prolonged segmental analgesia during the post-operative period with minimal systemic side effects (4). Pruritus and myoclonus are rarely observed side effects of neuraxial anesthesia in dogs. Pruritus has been described in 0.8% of dogs administered epidural morphine and/or bupivacaine (5). Side effects are more commonly observed after intrathecal opioid administration than after epidural injection (3,4). Myoclonus has been reported in 3 dogs following intrathecal administration of morphine in doses ranging from 0.15 to 2.3 mg/kg (6–8).

This case report describes the treatment of pruritus and myoclonus in a dog after subarachnoid administration of a combination of lidocaine, bupivacaine, and morphine.

Case description

A 6-year-old spayed female mixed breed dog, weighing 32 kg was scheduled for a tibial plateau levelling osteotomy. Pre-anesthetic medication consisted of acepromazine (Vanastress; Vana GmbH, Vienna, Austria), 0.02 mg/kg body weight (BW) and methadone (Heptadon; EBEWE Pharma GesmbH, Unterach, Austria), 0.1 mg/kg BW, intravenously (IV). Anesthesia was induced with propofol IV to effect (Propofol-Fresenius; FreseniusKabi GmbH, Bad Homburg, Germany) (total dose 4 mg/kg BW). The trachea was intubated and the endotracheal tube connected to an anesthetic circle system. Anesthesia was maintained with isoflurane (Isoflo, Abbot) in 100% oxygen, and an epidural injection of lidocaine (Xylanaestpurum 2%; Gebro Pharma GmbH, Fieberbrunn, Austria), bupivacaine (Carbostesin 0.5%; AstraZeneca GmbH, Wedel, Germany), and morphine (10 mg/mL, Vendal, Lannacher Heilmittel, Lannach, Austria) was planned as part of a balanced anesthetic protocol. All drugs for epidural injection were drawn from previously unused vials and the dog was positioned in sternal recumbency with the hind legs extended cranially. The skin over the lumbosacral space was clipped and aseptically prepared and a 20-G spinal needle (Yale; Becton Dickinson, Fraklin Lakes, New Jersey, USA) was inserted until a “pop” was felt when penetrating the ligamentum flavum. Free flow of cerebrospinal fluid (CSF) was observed on removal of the stylet indicating subarachnoid puncture. As a result, half of the intended volume was injected, resulting in a total dose of 0.05 mg/kg BW preservative-free morphine, 0.62 mg/kg BW lidocaine, and 0.31 mg/kg BW bupivacaine (total volume 3.16 mL). Anesthesia lasted 3 h and was uneventful with all measured cardiovascular and respiratory parameters within physiologic ranges. Lactated Ringer’s solution was infused at 10 mL/kg BW per hour throughout the procedure.

The dog started to lick and bite bilaterally at the base of the tail region about 15 min after extubation. The dog was awake and responsive, but had not regained motor function of her pelvic limbs. Palpation of the wound did not elicit a reaction; however, in response to tactile stimulation at the base of the tail violent biting of the region was elicited. As the dog became more alert, the licking increased in severity and occasional twitches of the tail were observed (about every 10 s). Acepromazine 0.02 mg/kg BW was administered IV as the dog became distressed after excluding a distended bladder or the need for defecation by abdominal palpation. Additionally, methadone, 0.1 mg/kg BW, IV and then ketamine (Ketasol Narketan10%; Vétoquinol Austria GmbH, Vienna, Austria), 0.5 mg/kg BW, IM were given to ensure sufficient analgesia. There was a decrease in licking, but not in muscular twitches, which became more severe over the 3 h following extubation. Rhythmic twitching (1 twitch per 10 s) affected the whole caudal half of the body and this progressed to clusters of approximately 10 twitches (1 per second) every 15 to 20 s, leading to a diagnosis of myoclonus. The dog became increasingly agitated and diazepam (Valium; Roche Austria GmbH, Vienna, Austria), 1 mg/kg BW was administered IV. Heavy sedation resulted and the frequency and intensity of myoclonus decreased. However, an additional injection of 0.5 mg/kg BW diazepam was necessary after 15 min due to recurrence of severe and frequent myoclonus. Buprenorphine (Temgesic; AESCA GesmbH, Traiskirchen, Austria), 0.01 mg/kg BW and butorphanol (Butomidor, Richter Pharma AG, Wels, Austria), 0.2 mg/kg BW were given IV 2 h and 3 h after extubation, respectively, with the intention of partially reversing the actions of morphine at the mu (MOP) receptor. Both agents proved to be ineffective in eliminating pruritus and myoclonus completely; however, signs of distress were not evident. The dog slowly recovered from sedation and 5 h after extubation showed infrequent episodes of pruritus (every 5 or 6 min) and less frequent myoclonus (1 twitch every 30 to 40 s). Frequency and severity of myoclonus and pruritus decreased further over time. At about 8 h after extubation they were only visible after tactile stimulation of the sacral region. All symptoms ceased about 10 h after extubation. The dog urinated normally at this time. Clinical examination performed the following day was unremarkable and the dog was discharged. The dog was lost to follow-up.

Discussion

This report describes pruritus and myoclonus after spinal anesthesia. One reason for an adverse drug reaction such as this, is the use of preservatives in neuraxially administered solutions. None of the drugs used contained preservatives and the data sheets of all agents indicate their suitability for neuraxial use in humans. Another reason for an adverse drug reaction is interaction between components of the injectate. Mixing local anesthetics and opioids is common practice to achieve excellent intraoperative analgesia (local anesthetic) combined with prolonged postoperative pain relief (opioid). Neural blockade produced by mixing local anesthetics is deemed unpredictable and controversial and depends on a number of factors (9). In this case, lidocaine and bupivacaine were mixed based on the perceived advantage of combining agents to achieve a quick onset and long duration of action (10). Retrospectively, the physicochemical properties (color, formation of precipitate, pH, specific gravity) of the individual solutions and of the mixture were assessed in a new set of bottles (Table 1). Lidocaine and bupivacaine are stable in pH 6 to 7, if used soon after mixing (9). Other authors have stated that pH may be unimportant when solutions are used for single administration since the buffering capacity in the body is sufficient to increase the pH of the anesthetic solution into the physiologic range (11). We assume that pruritus and myoclonus were a side effect of spinal anesthesia rather than from physicochemical changes due to mixing the drugs. Initial treatment of pruritus and myoclonus utilized conventional therapy: tranquilizers (acepromazine and diazepam), opioids (methadone, buprenorphine, and butorphanol), and an N-methyl-D-aspartate receptor antagonist (ketamine). These treatments were met with varying success and the waning effect of morphine over time is considered the main reason for improvement.

Table 1.

Physicochemical properties of lidocaine, bupivacaine and a lidocaine-bupivacaine mixture

Agent	pH	SG	Precipitate	Color
Lidocaine 2%	6.2	1.015	no	clear
Bupivacaine 0.5%	5.0	1.008	no	clear
Lidocaine 2%-bupivacaine 0.5% mixture (1:2)	5.5	1.011	no	clear

SG — specific gravity.

The occurrence of pruritus after neuraxial anesthesia was reported in 2/242 dogs (0.8%) after epidural morphine with or without bupivacaine (5). However, the publication does not detail if the dogs with pruritus received only morphine or both agents (5). Pruritus following intrathecal injection of morphine (0.1 mg/kg BW) with preservatives has been reported in sheep with 2/37 animals (5.4%) affected (12). In human patients receiving epidural or intrathecal opioids the incidence of pruritus has been reported as 8.5% and 46%, respectively (13). Pruritus in the dog described here was so severe that intervention was necessary as the dog became distressed while trying to lick and chew the affected area. Self-mutilation of the tail and pelvic limbs has been described after spinal morphine administration in dogs (3). As well as the unpleasant sensation of pruritus, the dog started to show myoclonus, which progressed over the first 3 h after extubation. Myoclonus after epidural morphine injection has been reported to be very rare (4). Two case reports of myoclonus in dogs after intrathecal morphine injection can be found in the literature. These dogs received 3 times and 30 times the intrathecal dose of preservative-free morphine reported here (6,7). Both dogs had to be managed with general anesthesia in order to control myoclonus; therefore, it is impossible to say if pruritus was present in these dogs.

This is the first description of a dog suffering from both pruritus and myoclonus after intrathecal injection of a commonly used dose of preservative-free morphine. Considering the pathophysiology of each of the symptoms it becomes evident that several mechanisms have to be involved. Itch is an unpleasant sensation stemming probably from afferent C fibers with excessive terminal branching eliciting a reflex or desire to scratch (14). With neuraxial administration peripheral mechanisms, like histamine release, causing itch are less likely (14). Several central mechanisms are proposed in the literature to be responsible for itching after administration of opioids; however, the exact mechanisms remain unclear. Opioid receptors in the superficial and deep dorsal horn neurons may be involved in signalling this sensation. It has been hypothesized that opioids are directly and indirectly involved in the facilitation of protective reflexes and itch and hyperalgesia are the manifestation of this. Pruritus may be caused by a facilitation of superficial neuronal response to histamine coupled with inhibition of deep dorsal horn neurons with intrathecal morphine administration (13). However, other receptors in the spinal cord or brain that may be involved include dopamine D2 receptors, 5-HT3 receptors, prostaglandin system, as well as GABA, and/or glycine receptors (14). In animals, similarities between hyperalgesia induced by intrathecal opioids and by glycine antagonism have been described (13). This hyperalgesia syndrome is not abolished by naloxone, contrary to the findings in humans in whom naloxone sometimes can reduce hyperalgesia and itch (13).

The pathophysiologic mechanism of myoclonus is not known. Studies in animals and spinal neuronal cultures have led to various theories explaining muscular hyperactivity: an interaction of morphine with non-opioid (GABA and glycine) receptors in the central nervous system, potentially blocking the postsynaptic inhibition (13); activation of spinal serotonergic systems causing myoclonic activity following high doses of intrathecal morphine (15); or indirect activation of N-methyl-D-aspartate receptors (has been implicated in causing myoclonus, but this has been questioned) (16). Evidence that non-opioid receptors are likely to be involved in the occurrence of myoclonus is that symptoms cannot be treated with naloxone (13,16).

The treatment of the dog herein followed the main recommendations found in the literature. Initially we administered acepromazine, as an emergence reaction from anesthesia was assumed, and the duration of action is 6 to 8 h. Distress due to a distended bladder or the need to defecate with incomplete motor control was ruled out by abdominal palpation. Severe acute pain causing licking and biting as a form of self-mutilation was a differential diagnosis to pruritus; however, several factors make this interpretation unlikely. There was minimal motor control of the hind limbs on recovery, indicating residual action of the local anesthetics, and minimal change of the symptoms with the administration of methadone and ketamine, making the sensation of pain unlikely. At this time the dog was responsive to its environment and it became obvious that the dog suffered from severe pruritus.

No clear guidelines exist for treatment of opioid-induced pruritus. For small animals, suggested treatment options include systemic antihistamines such as diphenhydramine and MOP receptor antagonists such as butorphanol, nalbuphine, and naloxone as well as acepromazine to eliminate anxiety until the epidural opioids have worn off (3). Acepromazine may have an additional benefit, as it acts on central dopamine and 5-HT3 receptors. As histamine release has been implicated to play a role by some authors, treatment with H1 and H2 blockers might be warranted (17). However, other authors questioned their usefulness and antihistaminic agents were not administered in this case (14).

Butorphanol [kappa (KOP) agonist- MOP antagonist] is effective in attenuating neuraxial morphine-induced itch without reducing morphine analgesia in primates (18). The use of buprenorphine, a partial MOP agonist for the treatment of non-opioid induced pruritus is described in humans (19). In the present study, butorphanol and buprenorphine were used in an attempt to reduce itching while maintaining an acceptable analgesic level after major surgery. Buprenorphine was given as it has a long duration of action; when, after 1 h no effect was obvious butorphanol was considered as the next option. However, neither butorphanol nor buprenorphine administration decreased the severity of pruritus in this case. Naloxone was not administered as a first line of treatment for ethical reasons, as inadequate analgesia might have resulted and the treatment with a KOP agonist- MOP antagonist may have been more suitable (19). The severity of the signs decreased and the dog was not distressed by the time the effects of buprenorphine and butorphanol were evaluated and administration of naloxone was not deemed necessary. Naloxone is effective in preventing or treating intrathecal or epidural opioid-induced itching in humans (13). However, there is some indication that naloxone may not be as useful in abolishing pruritus in animals as in humans (13).

The literature was subsequently reviewed and alternative treatment options specifically for the treatment of opioid-induced pruritus have been identified. Propofol depresses posterior horn transmission in the spinal cord, possibly reducing itching (14). It has been used at subanesthetic doses to treat opioid-induced pruritus; however, mixed results have been reported (19). Another therapeutic option would have been ondansetron, which resulted in variable results ranging from helpful to ineffective in humans (19). Dopamine D2 receptor antagonists (reduced intensity only) and droperidol have been used successfully in the treatment of opioid-induced pruritus in humans (19).

In the dog reported here, myoclonus occurred after recovery about 3 h after intrathecal injection, starting at the tail. Onset of generalized myoclonus after administration of 2.3 mg/kg BW morphine injected into the cisternal space is described after 15 min in an awake dog, resulting in death 50 min after administration (8). This report mentions that myoclonus starts in the pelvic limbs when morphine is administered in the lumbar intrathecal space (8). Other reports in anesthetized animals describe onset times of 50 and 90 min after lumbar intrathecal injection of 1.3 and 0.15 mg/kg BW of morphine, respectively (6,7). Two of the reports also describe worsening of signs with mechanical or tactile stimulation, similar to our description (7,8). Various treatment approaches for myoclonus have been described. Variable effects have been reported with the systemic or epidural administration of benzodiazepines and baclofen (6,7). The administration of diazepam brought temporary relief in the dog reported here and an additional dose had to be given, as the effect seemed short-lasting. The dog was also heavily sedated and by the time the dog recovered it seemed to be less distressed; therefore, further administration of benzodiazepines did not seem necessary. Benzodiazepines were also initially effective in the treatment of myoclonus in 2 aforementioned case reports in dogs (6,7).

Another option is the use of non-depolarizing muscle relaxants in anesthetized animals. Variable success has been reported with the use of atracurium in dogs (6,7). As the use of non-depolarizing muscle relaxants makes unconsciousness and intermittent positive pressure ventilation necessary we did not consider their use in the first instance. Pentobarbital was eventually used to control the myoclonus in one of the aforementioned dogs, while in the dog with the intrathecal morphine overdose it was ineffective (6,7). Pentobarbital was not necessary, but was discussed as the next treatment option in the dog reported here. As discussed previously, intravenous naloxone shows little effect on myoclonus after epidural or systemic injection as the pathophysiologic mechanism is not opioid receptor related.

Local anesthetics have been suggested to reduce opioid-induced myoclonus (21). One possible explanation for the relatively late onset of myoclonus compared with the other cases reported may be that intrathecal local anesthetics were administered together with morphine in our case (6–8). However, a recent case report has suggested intrathecal bupivacaine as the cause of myoclonus (21). Alternatively, the late appearance of myoclonus in our case might be due to the fact that ketamine, an N-methyl-D-aspartate receptor antagonist, was injected when pruritus became evident in order to achieve better non-opioid related analgesia. The administration of an N-methyl-D-aspartate receptor blocker might have delayed the appearance of myoclonus in relation to the pruritus. Indeed ketamine has been used to prevent morphine-induced hind limb myoclonic seizures in mice (22).

The time course and duration of side effects in the canine reports is very variable and in the dog reported here myoclonus and pruritus were absent at 13 h after intrathecal injection. Myoclonus was absent 9 and 22 h after administration of morphine in the 2 surviving dogs previously reported, although no earlier attempts to wean them off their respective treatment protocols had been made (6,7). The duration of side effects may depend on the duration of action of intrathecal morphine as well as the amount administered. Experimentally 0.01 to 0.04 mg/kg BW of intrathecal morphine provided an antinociceptive effect for 8 h, at which time data acquisition was terminated (23). Clinically, intrathecal use of morphine is mainly reported when puncturing the dura with a failed epidural approach. Current dose recommendations are 0.05 mg/kg BW for inadvertent dural puncture and 0.03 mg/kg BW for canine spinal surgery patients (3,4,24).