Equine Veterinary Journal ISSN 0425-1644 DOI: 10.1111/evj.12265
Grayson-Jockey Club Research Foundation: Review Article
Pain control in horses: What do we really know? L. C. SANCHEZ* and S. A. ROBERTSON† Department of Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, USA † Department of Small Animal Clinical Sciences, College of Veterinary Medicine, Michigan State University, East Lansing, USA. *Correspondence email: [email protected]
ﬂ.edu; Received: 07.12.13; Accepted: 07.03.14
Summary Currently, approaches to pain control in horses are a mixture of art and science. Recognition of overt pain behaviours, such as rolling, kicking at the abdomen, ﬂank watching, lameness or blepharospasm, may be obvious; subtle signs of pain can include changes in facial expression or head position, location in the stall and response to palpation or human interaction. Nonsteroidal anti-inﬂammatory drugs (i.e. phenylbutazone, ﬂunixin meglumine and ﬁrocoxib), opioids (i.e. butorphanol, morphine and buprenorphine) and α2-adrenergic agonists (i.e. xylazine, detomidine, romiﬁdine and medetomidine) are the most commonly used therapeutic options. Multimodal therapy using constant-rate infusions of lidocaine, ketamine and/or butorphanol has gained popularity for severe pain in hospitalised cases. Drugs targeting neuropathic pain, such as gabapentin, are increasingly used for conditions such as laminitis. Optimal strategies for management of pain are based upon severity and chronicity, including special considerations for use of intra-articular or epidural delivery and therapy in foals. Strategies that aim to mitigate adverse effects associated with use of various analgesic agents are brieﬂy discussed. Keywords: horse; analgesia; chronic pain; acute pain
Introduction Currently, approaches to pain control in horses lack a robust evidence base. Although research reports of antinociceptive and analgesic therapy in horses have certainly gained ground over the years, most involve models of healthy horses or retrospective evaluation of clinical cases. Reports of prospective clinical trials are few and far between; thus, most practitioners base analgesic choices upon a combination of the available literature and clinical experience. While colic and lameness are common sources of visceral and somatic pain, respectively, other sources of pain in equine cases include ophthalmic diseases and surgical manipulation. Recognition of overt pain behaviours, such as rolling, kicking at the abdomen, ﬂank watching, lameness or blepharospasm, may be obvious; subtle signs of pain can include changes in facial expression or head position, location in the stall and response to palpation or human interaction. This review presents positive and negative aspects of drug options available in commonly encountered clinical situations. Recommended dosages are given in Tables 1 and 2.
Management of acute, moderate and perioperative pain For the purposes of this review, examples of horses with acute moderate pain include those with a nonstrangulating obstruction of the large colon, minor foot abscess or uncomplicated corneal ulceration. In many circumstances, nonsteroidal anti-inﬂammatory therapy is sufﬁcient. Clinically, perioperative pain can vary widely from that associated with castration to much more severe clinical situations, such as surgery for intestinal resection or fractures. The latter situations frequently necessitate multimodal therapy, which will be discussed separately. N-Butylscopolammonium bromide (NBB) has both anticholinergic and antispasmodic properties for a very short duration (15–20 min) and is Equine Veterinary Journal •• (2014) 516–522 © 2014 EVJ Ltd
frequently an excellent choice for the treatment of nonstrangulating obstructions of the large colon, including spasmodic colic . Administration of NBB has demonstrated antinociceptive effect in models of caecal [2,3] and colorectal distension . The administration of NBB also decreases rectal tone and facilitates rectal examination . Given that NBB administration, either alone or in combination with xylazine, causes tachycardia and alterations in blood pressure , care should be taken with its administration to animals with systemic compromise or pre-existing moderate to severe tachycardia. Nonsteroidal anti-inﬂammatory drugs (NSAIDs) are probably the most frequently used analgesic agents in horses worldwide, primarily because many of the commonly occurring causes of pain are mediated by inﬂammation. Thorough reviews of these drugs are presented elsewhere . Flunixin meglumine and phenylbutazone are most commonly used for colic and lameness, respectively, but are similarly effective for the latter . Of note, experimental work has shown that the practice of ‘stacking’ (administration of ﬂunixin and phenylbutazone in combination) did not improve analgesia in a lameness model , but co-administration of phenylbutazone (2.2 mg/kg bwt per os every 12 h) and ﬂunixin (1.1 mg/kg bwt i.v. every 12 h) did improve lameness in clinical cases beyond phenylbutazone alone . Also, decreasing doses (0.5 vs. 1 mg/kg bwt ﬂunixin) can limit analgesic effect in a lameness model , whereas increasing dosage (8.8 vs. 4.4 mg/kg bwt/day phenylbutazone) did not provide additional analgesic effects in clinically lame horses . Agents demonstrating selectivity for cyclo-oxygenase-2 in the horse, such as ﬁrocoxib, have demonstrated similar analgesic efﬁcacy to phenylbutazone in horses with naturally occurring lameness [13,14]. The α2-adrenergic agonists are an excellent choice for sedation and short-term analgesia. Thorough reviews of their actions are available elsewhere . It is important to note that sedative and analgesic effects may not be of equal duration for the majority of drugs in this class [16,17]. The analgesic effects are primarily mediated through spinal and supraspinal actions; thus, epidural administration of these drugs can be particularly effective; co-localisation in the spinal cord with some opioid
L. C. Sanchez and S. A. Robertson
TABLE 1: Recommended doses for commonly used analgesic medications Dosage (mg/kg bwt unless noted)
Nonsteroidal anti-inﬂammatory drugs
Flunixin Phenylbutazone Firocoxib Meloxicam Xylazine Detomidine Medetomidine Butorphanol
Ketamine Lidocaine N-Butylscopolammonium bromide
0.5–1.1 2.2–4.4 0.1 0.6 0.2–1.1 0.005–0.04 0.004–0.01 0.01–0.05 0.04–0.1 18 μg/kg bwt, then 10–23 μg/kg bwt/h 0.0075–0.01
0.4–1.2 mg/kg bwt/h 1.3 mg/kg bwt bolus, then 3 mg/kg bwt/h 0.3
receptors can contribute to a synergistic effect when α2-agonists and opioids are administered simultaneously [18,19]. Xylazine provides excellent dose-dependent visceral analgesia for short durations, typically 15–20 min, but up to 60 min in some pain models [20–22]. Somatic antinociception of very short duration is also provided . Visceral antinociception after detomidine administration has been demonstrated in caecal , duodenal and colorectal distension models  for slightly longer periods (90 min). Romiﬁdine can provide longer periods of somatic antinociception (120 min) . Medetomidine can be used as part of a balanced anaesthetic protocol , including combination with morphine for standing laparoscopy in horses . Opioids are commonly used in conjunction with an α2-adrenergic agonist for either single or repeated injections or as part of a perianaesthetic protocol. Sedatives are commonly administered in conjunction with opioids to conscious horses in order to diminish the potential for central nervous system excitation. Butorphanol is probably the most frequently used opioid, especially for visceral pain . Recent studies using a thermal threshold model suggest that buprenorphine
i.v., per os i.v., per os i.v., per os per os i.v., i.m. i.v., i.m. i.v. i.v. i.m. i.v.
12–24 12–24 24 24
Sedation outlasts analgesia
3–4 4–6 Constant-rate infusion
Constant-rate infusion Constant-rate infusion
Can combine with α2-agonists
Co-administration of acepromazine appears to minimise excitement-related potential adverse effects Co-administration of sedatives can minimise excitement, especially at higher doses 75 ml/h for 500 kg horse (2%)
provides superior analgesia relative to butorphanol  and for perioperative pain associated with castration . Co-administration of acepromazine appears to minimise excitement-related potential adverse effects [28,30]. In research settings, morphine (0.05 and 0.1 mg/kg bwt i.v. or i.m.) had no antinociceptive (thermal or electrical) effect . Clinically, morphine (0.1–0.2 mg/kg bwt i.v.) signiﬁcantly improved the quality of recovery when added to a standard anaesthetic protocol in horses undergoing upper airway surgery in one report , whereas horses anaesthetised for elective procedures had better recovery scores and fewer alterations in anaesthetic depth when receiving an intraoperative constant-rate infusion of dexmedetomidine (1.75 μg/kg bwt/h) relative to morphine (0.1 mg/kg bwt/h) . Two other studies reported favourable results with perioperative morphine use [34,35]. Perioperative morphine use may predispose horses to post operative colic, but results have been somewhat contradictory, showing a 4-fold risk of colic following orthopaedic surgery in one report , but in another retrospective report of horses anaesthetised for magnetic resonance imaging or nonabdominal surgery, colic was associated with surgery but not
TABLE 2: Dosages of drugs used alone (the ﬁrst 5 drugs) or in combination (the next 4 sets) for caudal epidural analgesia in horses Drug
Dosage (mg/kg bwt)
Lidocaine Bupivacaine Xylazine Detomidine Morphine
0.2 mg/kg bwt 0.04–0.06 mg/kg bwt 0.03–0.35 0.02–0.06 0.05–0.2
0.5–1.5 3.5–5 1–2 2–4 3–8
5–8 ml 5–8 ml; concentration of 0.25–0.5%; 0.25% preferred Typically 0.2 mg/kg bwt
0.22 mg/kg bwt 0.17 mg/kg bwt
Xylazine Morphine Detomidine Morphine
Dilute with saline to total volume of 30 ml; may cause pruritis
Equine Veterinary Journal •• (2014) 516–522 © 2014 EVJ Ltd
L. C. Sanchez and S. A. Robertson
morphine administration . A recent preliminary study documented that morphine-3-glucuronide, a metabolite associated with neuro-excitation in mice, was present in substantially higher concentrations after i.v. bolus administration of morphine to horses than was the metabolite morphine-6-glucuronide . Behavioural changes noted following morphine administration included sweating and muscle fasiculations after a dose of 0.2 mg/kg bwt, and ﬂared nostrils, muscle tremors and ataxia after administration of 0.5 mg/kg bwt; behavioural changes were not evident after administration of either 0.05 or 0.1 mg/kg bwt . Methadone has effects at μ-opioid and N-methyl-D-aspartate receptors  and, when given i.v. or per os (0.15 mg/kg bwt), resulted in plasma concentrations known to be analgesic in other species with no apparent adverse effects; intragastric administration resulted in much lower bioavailability than oral administration [40,41]. Levomethadone has been shown to increase and prolong the antinociceptive effects of detomidine .
Management of acute, severe pain: multimodal therapy Given that horses in severe pain can have either central sensitisation and/or stress-induced hyperalgesia, combining analgesic agents with different mechanisms of action can be of beneﬁt. In the authors’ experience, multimodal therapy in an acute, hospital situation typically involves NSAID administration in combination with a constant-rate infusion of lidocaine, ketamine, butorphanol or a combination thereof. In one prospective clinical trial, horses receiving butorphanol constant-rate infusion (13 μg/kg bwt/h) in addition to ﬂunixin after colic surgery lost signiﬁcantly less weight, had improved recovery characteristics and on average were discharged 3 days earlier than a similar population of horses receiving ﬂunixin and a saline constant-rate infusion . Lidocaine, an aminoamide local anaesthetic, prevents propagation of action potentials by binding to voltage-gated sodium channels and has potential analgesic, prokinetic and anti-inﬂammatory properties [44–49]. Clinical signs of toxicity in conscious horses include skeletal muscle tremors, altered visual function, anxiety, ataxia and collapse. Although few reports detail an analgesic effect in clinical cases, i.v. lidocaine infusion has been shown to obtund the electroencephalographic response to castration in anaesthetised ponies  and increase the nociceptive threshold in response to a thermal stimulus in conscious horses . Some clinicians ﬁnd lidocaine particularly useful in the post operative period following exploratory laparotomy, although effects may be related to anti-inﬂammatory or prokinetic, rather than direct analgesic, properties. Lidocaine infusion has been associated with reduced small intestinal size and peritoneal ﬂuid accumulation  and improved survival  in this setting. Intraoperative use of lidocaine was thought to reduce the incidence of post operative ileus by ∼50% , and perioperative infusion decreased the volume and duration of reﬂux in comparison with saline-treated control animals in a multicentre study of horses with enteritis or post operative ileus . As clinical cases receiving lidocaine alone or in combination often require long-term therapy, one should be careful to monitor dose and desired effect over time, especially when used in conjunction with highly protein-bound drugs, such as ceftiofur sodium and ﬂunixin meglumine , general anaesthesia or in horses which are otherwise systemically compromised. Although drug accumulation was not noted in healthy horses over a 96 h study period , it has been demonstrated in a clinical setting . Higher serum concentrations have been achieved in anaesthetised compared with awake horses , due to decreased cardiac output and hepatic blood ﬂow and clearance, and can reach a range reported to be toxic in conscious horses . Ketamine, a noncompetitive N-methyl-D-aspartate receptor antagonist, can modulate central sensitisation and exert an antihyperalgesic effect at subanaesthetic doses [57,58]. Although the majority of work documenting these effects has been performed in other species, ketamine is likely to be most useful in situations where pain is severe or where an individual’s pain threshold has been altered such that a previously nonpainful stimulus becomes painful. Antinociception has not been demonstrated in healthy Equine Veterinary Journal •• (2014) 516–522 © 2014 EVJ Ltd
horses receiving infusions of ketamine , but beneﬁcial effects have been noted in clinical settings [60,61]. A combination of xylazine, butorphanol and ketamine (0.2 mg/kg bwt i.v.), relative to xylazine alone or xylazine/butorphanol, did not increase the level of sedation but improved ease of insertion of a dental ﬂoat, decreased response associated with arthrocentesis and increased pressure algometry threshold at the withers, but increased responsiveness to a sharp needle prick . In a small clinical report, transdermal fentanyl appeared clinically effective in horses with pain refractory to NSAID therapy, especially in animals with a lower body weight . This mirrors the clinical impression of some clinicians that transdermal fentanyl may provide good clinical analgesia in foals, but is contrary to research reports in adult horses demonstrating that serum concentrations high enough to provide either a reduction in the minimum alveolar concentration of isoﬂurane  or visceral or somatic antinociception  can be associated with adverse effects, such as hyper-responsiveness. Uptake of fentanyl from a transdermal patch is highly variable in adult horses and foals; variation has been reported between individuals and in association with the location of patch placement [65–67]. Thus, although transdermal fentanyl is convenient and easy to use over clipped hair in horses, evidence supporting its efﬁcacy for analgesia in adult horses is currently lacking.
Management of chronic pain Chronic orthopaedic disorders, including osteoarthritis, are frequently encountered in clinical practice. In many cases, NSAIDs provide some degree of relief, but long-term administration is required. In these situations, use of agents with increased cyclo-oxygenase-2 selectivity, such as ﬁrocoxib, may decrease the potential for adverse effects associated with NSAID administration . Manual therapies, such as massage, may be useful in such situations, but more data are needed regarding their efﬁcacy; a thorough review is available elsewhere . Chronic laminitis is one of the most challenging problems facing the equine veterinarian, especially as many horses are refractory to NSAID therapy. A neuropathic component has been demonstrated, making ketamine and gabapentin suitable candidate drugs . In one report, oral tramadol (5 mg/kg bwt every 12 h) alone provided little pain relief, but the combination of tramadol and ketamine (0.6 mg/kg bwt/h i.v. for 6 h for the ﬁrst 3 days of a 7 day treatment) resulted in decreased blood pressure, decreased forelimb ‘ofﬂoading’ frequency and increased forelimb load in a randomised crossover study of horses with naturally occurring laminitis . In other reports, tramadol appears to have very low oral bioavailability (∼3%) and a short half-life  and, when given i.v. (0.1–1.6 mg/kg bwt), did not alter hoof withdrawal or skin-twitch latency to a thermal stimulus . Thus, current evidence does not support the use of oral tramadol alone in horses. Gabapentin and lidocaine have provided analgesia in a rat model of neuropathic pain , and gabapentin administration reportedly improved hindlimb pain that was probably associated with femoral neuropathy in one horse . Gabapentin has also been suggested to reduce head-shaking behaviour, which may have a component of neuropathic pain . Gabapentin has a relatively low bioavailability, but no apparent adverse effects following oral administration in horses [75,76]. Further work is needed to assess the clinical effect of gabapentin more objectively in horses with clinical pain. Newer therapies, such as soluble epoxide hydrolase inhibitors and vanilloid receptor antagonists, may prove useful, but further work is needed [77–79].
Management of severe pain: epidural and intra-articular drug administration Caudal epidural injection can provide analgesic effects with signiﬁcantly decreased systemic adverse effects in horses with severely painful hindlimb conditions. If repeated injections are required, an epidural catheter can be placed for long-term treatment; detailed instructions regarding placement are available elsewhere . Local anaesthetic agents produce loss of both sensory and motor function; thus, the volume administered must be calculated carefully to
avoid cranial spread, which can result in recumbency. Lidocaine (2%) is commonly used for short-term management; anaesthesia usually occurs within 5 min and lasts 45–60 min. Recommended doses range from 0.2 to 0.25 mg/kg bwt (1.0–1.25 ml per 100 kg bwt). Use of opioids, tramadol, α2-adrenoceptor agonists and ketamine allows maintenance of motor function but may induce mild ataxia. Used individually, these agents provide limited analgesia; when used in combination, their analgesic effects can be profound. Morphine (0.15–0.2 mg/kg bwt) has a slow onset (1–5 h) but long duration (from 16 to 28 h) of action, and has been shown to decrease lameness, improve weightbearing at rest and range of motion during locomotion in an acute synovitis model . Tramadol has a somewhat slow onset (∼30 min), with a reasonable duration (5 h) of action . Butorphanol  and methadone  can also be used as sole agents or in combination. Xylazine and detomidine [84–87] both produce sensory blockade, but the duration is longer with xylazine, with fewer systemic effects . Ketamine provides dose-dependent analgesia for a relatively short duration (30–75 min) . Combination therapy, often including either an α2-adrenergic agonist (e.g. detomidine 0.02–0.03 mg/kg bwt) or lidocaine with morphine (0.1–0.2 mg/kg bwt) [89,90], is most relevant clinically. Long-term catheterisation can be associated with mild inﬂammatory changes in cerebrospinal ﬂuid, but clinically signiﬁcant adverse events are rare [91,92]. Sedation often occurs soon after epidural injection of α2-adrenergic agonists but is short lived compared with analgesia. Ataxia and recumbency can be prevented by avoiding oversedation, injecting slowly, limiting the volume of local anaesthetic injected and using caution in weak or debilitated horses. Pruritus has been reported after epidural administration of morphine in horses , as it has in other species. A variety of agents can be added within synovial structures as part of a multimodal approach for relief of severe orthopaedic pain. Intra-articular blocks may be best performed with mepivacaine, because the severity of toxic effects of this local anaesthetic on equine articular chondrocytes is less than that for lidocaine and markedly less than that for bupivacaine . In an in vivo model using healthy horses, single intra-articular injections of bupivacaine or lidocaine increased biomarkers of collagen synthesis . Mu-opioid receptors have been identiﬁed in equine synovial tissue , and some clinicians advocate intra-articular administration of opioids to relieve pain related to arthroscopic surgery. Morphine alone or combined with ropivacaine produced a good analgesic effect for up to 24 h in an experimental synovitis model, whereas ropivacaine alone provided pain relief of only limited effect and duration (