Gastrointestinal Ileus

Chris Sanchez • Guy D. Lester • Samuel L. Jones

A number of factors are responsible for movement of feed through the gastrointestinal tract, and disturbance in any of these can result in ineffective transit and signs of abdominal pain.

Several equine diseases and syndromes that probably involve altered gastrointestinal motility include POI, cecal emptying defect, gastroduodenal ulcer disease, simple bowel distention, intraluminal and extraluminal obstructions, peritonitis, and IBDs. Ineffective intestinal motility can also be a feature of several neonatal disorders, including prematurity, systemic sepsis, and neonatal encephalopathy, and can be found in postparturient mares. Certain parasitic infections, electrolyte derangements, and endotoxemia can modify intestinal transit in horses of all ages. Intestinal motility is also disturbed by general anesthesia and sedation, especially with the α₂ receptor agonists.

■ Physiology of Intestinal Motility The control of intestinal motility is complex and involves a combination of central innervation, autonomic innervation, and the enteric nervous system. Intestinal contractions are controlled primarily by the enteric nervous system and do not require extrinsic neural input. The inherent rhythmicity of electrical activity in the intestine is controlled by the interstitial cells of Cajal. These are highly specialized cells that are electrically coupled to the smooth muscle syncytium via gap junctions.¹ They are responsible not only for generation of slow waves (cyclical electrical activity) but also for coordination of pacemaker activity and propagation of slow waves along the intestine.

The enteric nervous system plays an important role in control and coordination of intestinal contraction. Contractile events are influenced by central and autonomic innervation, but external neural input is not required for contraction. The parasympathetic supply to the gastrointestinal tract is via the vagus and pelvic nerves, and the sympathetic supply is through postganglionic fibers of the cranial and caudal mesenteric plexuses. A complex network of interneurons within each plexus integrates and amplifies neural input, and the intensity and frequency of resultant smooth muscle contractions are pro­portional to sympathetic and parasympathetic input. Additional binding sites for a number of other endogenous chemicals, including dopamine, motilin, and serotonin, can be found within the enteric nervous system and on smooth muscle cells.⁵ Of importance is that mechanisms to slow progressive intestinal motility are also critical in order to retain feed for adequate digestion and absorption of nutrients. The terms jejunal and ileal brake have been used to describe the slowing of transit caused by mediators such as peptide YY, noradrenergic nerves, and opioid, serotonergic, and chemosensitive afferent neurons.⁶

Acetylcholine is the dominant excitatory neurotransmitter in the gastrointestinal tract and exerts its action through muscarinic type 2 (M₂) and and type 3 (M₃) receptors.

■ Mediators of Gastrointestinal Ileus and Major Disease Categories Inflammation within the intestinal muscularis layers and inhibitory neural events are important initiators of intestinal ileus.^14,15 Intestinal inflammation is important in primary inflammatory intestinal diseases in horses, such as DPJ and colitis, but it is also induced after simple intestinal handling during laparotomy. Experimental data from other species suggest that handling of the small or large intestine at the time of surgery induces an intense inflammatory response within the muscularis layers and a reduction in the intensity of smooth muscle contraction.^15-17 Furthermore, the inflamed intestine fails to contract normally in response to putative prokinetic agents.

The inhibition of motility associated with peritoneal inflam­mation may also be mediated through neural reflexes. The afferent segment is composed partly of capsaicin-sensitive visceral afferent C-fibers that terminate in the dorsal horn of the spinal cord, where they can activate inhibitory sympathetic fibers or, alternatively, synapse directly on the sympathetic ganglia. Consequently, the efferent limb of the neural pathway of the reflex expresses increased sympathetic outflow, mediated primarily through stimulation of α₂-adrenoreceptors, with subsequent inhibition of acetylcholine release. This provides the rationale for α₂ blockade in the treatment of ileus. Intra­luminal infusion of capsaicin before abdominal surgery ame­liorated the severity of POI in experimental rats.¹⁹ This finding highlights the importance of visceral afferent fibers in the development of POI.

Ileus can also occur with intestinal obstruction or displace­ment. Mild to moderate distention of the bowel, such as that occurring in the early stages of an intraluminal obstruction, evokes an increase in local contractile activity. ^, Excessive distention results in inhibition of motility within the distended segment of bowel. Intestinal stasis is not always detrimental and, in certain conditions, may be protective.

Endotoxemia is a feature of many diseases of the equine gastrointestinal tract, and endotoxins can independently exert a negative effect on intestinal motility and transit.²² Various mediators are probably involved, but activation of α₂- adrenoreceptors and production of prostanoids appear to be important, as the inhibitory effects of experimental endotoxin infusion are ameliorated by pretreatment with yohimbine or NSAIDs (phenylbutazone or flunixin), respectively.^23,24 Endo­toxin infusion induced an inflammatory response in the intestine of rats that mimicked that induced by handling during laparotomy.²⁵

■ Cecal Emptying Defect (Cecal Impaction) An important and specific motility problem involving the cecum occurs sporadically in adult horses and foals.^26-30 The pathophysi­ologic origin of cecal emptying defect is not known.

Retrospective data regarding cecal impactions suggest that two subtypes exist, and retrospective studies highlight those differences. In a United Kingdom report of 20 horses undergo­ing surgery because of cecal impaction, the condition appeared after orthopedic (10 of the 20) or colic (5 of the 20) surgery in the majority of horses.³³ Short-term survival in that report was 65%, and all but 3 horses underwent cecal bypass. Alter­natively, in a Texas report of 114 horses with cecal impaction, approximately half were treated medically and half surgically.³⁴ Of the horses allowed to recover from surgery, 95% survived to discharge, and typhlotomy alone was performed in 34 of 37 horses. Eighty-one percent of horses treated medically survived to discharge.³⁴ These reports appear to describe the two types of cecal impactions well, and the secondary (also called type II) impactions appeared to be associated with cecal emptying defect, and are therefore often categorized as a form of POI. In horses, postoperative cecal impaction was evident at a mean of 3.6 and 7.5 days after orthopedic and colic surgery, respectively.³³ The earliest detectable signs include lethargy and a reduction in appetite, fecal output, or both. Ineffective emptying results in overfilling of the cecum with moist contents, which is manifest by signs of mild to moderate colic.

■ Postoperative Ileus The most frequently reported form of ileus occurs after laparotomy and is termed simple or uncomplicated postoperative ileus. The condition in which intestinal motility is disturbed for longer periods after surgery, typically for more than 72 hours, is called complicated or paralytic ileus.³⁵ POI in horses is most commonly associated with surgery of the small intestine, particularly after resection and anastomosis. As stated previously, intestinal inflammation, particularly in the muscularis layers, is probably a key factor associated with the development of POI. The true incidence of POI in horses undergoing laparotomy for gastrointestinal disease is difficult to determine because of variation in definition, surgical skill, and anesthetic technique. POI was reported as a complication in only 10% of horses that underwent laparotomy for a small intestinal lesion at one referral practice³⁶ but in 27% in another.³⁷ In the latter study, horses with DPJ were included, but 30% of horses developed POI after small intestinal resection. In a multicenter study, 47 (19%) of 251 horses undergoing lapa­rotomy for acute colic, including those with small and large intestinal disease, developed POI.³⁸ As stated previously, the retrospectively collected POI data are extremely hard to interpret, and many mechanical obstructions and surgical complications can mimic POI³⁹; thus authors are encouraged to adopt uniform definitions of POI and make every attempt possible to distinguish POI from intestinal obstruction.⁴⁰

POI can have a significant negative effect on short-term postoperative survival.^41-43 Motility dysfunction is probably present in all horses after laparotomy, but most cases are subclinical and necessitate minimal intervention. In symptomatic animals, clinical signs are typically observed within 24 hours of recovery; these signs include colic, tachycardia, dehydration, decreased borborygmi and fecal output, and sequestration of fluid within the stomach. Rectal examination and ultrasonog­raphy may reveal small intestinal distention with rare or no wall movement. The severity and duration of intestinal stasis is variable, lasting from minutes to days.

The development of small intestinal POI, but not cecal emptying dysfunction, is influenced by the duration of 2738

surgery., technique may have a weak influence on small intestinal POI after jejunojejunostomy. The duration of intestinal ileus was shorter in animals that received a side-to-side stapled anastomosis than in those that had a hand-sewn end- to-end procedure.⁴¹ The duration of ileus after stapled end- to-end anastomosis was no different from that after either of the other procedures.

Reported risk factors for the development of POI in horses include age (>10 years), small intestinal resection and anasto­mosis, breed, heart rate at admission, large volume (>8 L) 37384344 reflux at admission, and duration of anesthesia and surgery. ’’’ Jejunocecostomy and side-to-side jejunojejunostomy have been associated with an increased risk of POI.^36,45 Pelvic flexure enterotomy and intraoperative and postoperative lidocaine infusion have been associated with a decreased incidence of _POI 38,43-45

■ Miscellaneous Motility Disruptions Other disruption of normal intestinal motility can occur in horses. Many, but not all, are associated with analgesic or sedative drug administra­tion. The inhibitory effects of α₂-adrenergic agonists, such as xylazine and detomidine, on intestinal motility in horses are well described.^46-54 The activation of postganglionic cholinergic nerves in the myenteric plexus by this class of drugs results in reduced release of acetylcholine from autonomic and enteric nerves. Intravenous xylazine inhibits cecal and large colon motility for 20 to 30 minutes without seriously disrupting small intestinal myoelectric activity, and detomidine can reduce large intestinal myoelectric activity for up to 3 hours. The α₂-antagonist yohimbine has a weak but positive effect on cecal emptying in normal ponies, which suggests that normal motility is under constant α₂-adrenergic tone.⁴⁸

Other analgesic medications can also decrease gastrointestinal motility. Opioids such as morphine have long been associated with decreased motility in humans.⁵⁵ Morphine has been shown to delay gastrointestinal transit and decrease gastrointestinal moisture content in horses.^56,57 Butorphanol administered alone appears to reset the antroduodenal MMC without substantial negative effects on motility, but it decreases duodenal motility when combined with xylazine more than does xylazine alone.^50,58 The roles of selective or preferential μ- and κ-receptor agonists vary different regions of the equine gastrointestinal tract: κ-Receptor agonists appear to have a more inhibitory effect in the jejunum, whereas μ-receptor agonists appear to inhibit the large colon more significantly.^59,60 Ketamine has also been shown to increase total gastrointestinal transit in healthy horses.⁶¹

Atropine is a postganglionic blocking agent that binds to muscarinic receptors. When administered at 0.04 mg/kg, individual small intestinal, cecal, and colonic contractions are inhibited for approximately 120 minutes, but small intestinal and colonic MMCs are suppressed for up to 8 hours.⁶² The negative effect of atropine on gastrointestinal motility in horses can also be evident with topical ophthalmic administration of atropine.⁶³ This effect, in combination with pain and other management factors, probably relates to the high reported risk of colic in horses hospitalized for ocular disease.⁶⁴

Gastric and small intestinal ileus, with no identifiable mechanical or physical disruption of motility, have also been identified as causes of acute colic in postparturient mares.⁶⁵ The origin of this condition has not been identified, but most mares responded well to decompression.

■ Diagnosis The diagnosis of ileus is based on history and physical examination findings. Important tests include deter­mination of pulse rate and rhythm, auscultation and percussion of the abdomen, rectal palpation, and passage of a nasogastric tube. CBC with fibrinogen estimation and cytologic analysis of peritoneal fluid may improve the accuracy of diagnosis. Affected animals may be colicky because of accumulation of fluid in the upper gastrointestinal tract (classical POI) or cecal contents (cecal emptying defect). Decompression of the stomach is important diagnostically and therapeutically in horses with POI after small intestinal surgery. The presence of intragastric fluid upon nasogastric intubation is a commonly applied diagnostic criterion, but little information is available to define the parameters that increase the predictive value of nasogastric reflux either alone or in combination with other clinical findings. Proposed criteria include gastric reflux of 4 L or more upon any single intubation or of more than 2 L/h on serial intuba­tions; gastric fluid pH higher than 4.0; persistent tachycardia (>40 beats/min); signs of mild to severe abdominal discomfort; and rectal or ultrasonographic evidence of multiple loops of fluid-distended small bowel.⁴⁰ Failure to relieve pain with gastric decompression could indicate mechanical obstruction, severe inflammation of the intestine, or peritonitis. Most animals with ileus are depressed and have reduced fecal output and intestinal borborygmi. Intestinal sounds should, however, be interpreted with caution, because the presence of borborygmi does not always equate to progressive intestinal motility and may merely reflect local, nonpropagated contractions. Cecal distention with either firm or soft contents can be palpated rectally in horses with cecal dysfunction.

It is important to distinguish functional ileus from mechanical obstruction. This can be extremely difficult, but horses with mechanical obstruction typically have sustained high volumes of gastric reflux that vary little over time. Distinguishing mechanical obstruction from POI is especially important because horses with mechanical obstruction in the postoperative period may benefit from early repeat laparotomy.³⁹

■ Treatment

SUPPORTIVE CARE. The management of gastrointestinal ileus is dependent on the segment of gastrointestinal tract involved. Therapy for ileus of the proximal gastrointestinal tract typically involves a combination of gastric decompression, fluid and electrolyte therapy, and antiinflammatory drugs. Nasogastric decompression is the mainstay of management to prevent excessive distention of the stomach and small intestine. This often necessitates placement of an indwelling nasogastric tube, a procedure that some authors have speculated may in itself delay gastric emptying. In one study, the placement of an indwelling tube for 18 hours did not adversely affect the emptying of liquids, but in another study, a delay was noted when the tube was left in place for 72 hours.^66,67 Both studies used apparently healthy animals.

Electrolyte balance is important, particularly with regard to maintaining adequate extracellular concentrations of potas­sium, calcium, and magnesium. Calculation of appropriate fluid replacement volume should include maintenance requirements (40 to 60 mL/kg/day) and an estimate of losses, especially nutrients lost through gastric decompression. Overhydration should be avoided to prevent intestinal edema. Intravenous colloidal fluids may be useful in addition to crystalloid fluids to prevent overhydration. Whereas medical therapy with fluid replacement and antiinflammatory drugs may be helpful for horses with primary cecal impaction, surgical therapy is probably required for horses with secondary cecal emptying defect.^33,34,68 Surgical management may include typhlotomy alone, typh­lotomy with a bypass procedure such as ileocolic or jejunocolic anastomosis, or a bypass without typhlotomy.^33,68 If affected horses survive postoperatively to discharge from the hospital, chances of long-term survival appear good.³³

Evidence in experimental models and in human medicine suggests that early refeeding after surgery helps prevent POI.⁶⁹ Early refeeding was found to be beneficial in equine surgical cases.³⁶ Many clinicians offer small (one or two handfuls) meals of good-quality roughage frequently to reduce the risk of POI. A significant clinical or economic benefit of parenteral nutrition in adult horses with gastrointestinal disease also has yet to be demonstrated.^70-72 Parenteral nutrition should be considered when feed has been withheld for more than 96 hours, particularly in horses with a surgical wound, increased metabolic demands, or risk of hyperlipidemia. Limited exercise, in the form of hand-walking, may provide some benefit to these animals, although there is no evidence that exercise has a direct effect on intestinal motility in either horses or humans with POI.⁷³

The pivotal role of intestinal inflammation provides a strong rationale for the use of antiinflammatory drugs in affected animals. Flunixin meglumine is widely used in equine practice as an analgesic and antiinflammatory agent, and ameliorates many of the adverse systemic effects of endotoxin. A potential negative effect of NSAIDs on large intestinal contractility has been suggested on the basis of results of in vitro studies⁷⁴; however, a similar response has not been reported in whole animal 464875

studies.⁴⁶,⁴⁸,'⁵ There is some concern that non-cyclooxygenase- selective drugs such as flunixin meglumine inhibit anastomotic site healing and mucosal repair. An alternative may be firocoxib, which is highly selective for COX-2 and is as effective as flunixin in controlling colic pain.⁷⁶ Lidocaine is also an alternative or adjunct analgesic and antiinflammatory drug. The use of lidocaine is discussed in more detail later in this chapter.

Broad-spectrum antimicrobial drugs are indicated when sepsis is suspected or if the immune system is compromised, as in cases of moderate to severe endotoxemia. Theoretic concerns have been raised regarding the use of aminoglycoside antibiot­ics in animals with ileus. Inhibition of intestinal contractions occurred when sections of intestine were exposed to high concentrations of aminoglycoside antimicrobial drugs, but this inhibitory effect is unlikely to occur at clinically relevant doses.⁷⁷ The administration of benzylpenicillin or ceftiofur was associated with an increased risk of colic in the week after anesthesia for diagnostic imaging or nonabdominal surgical procedures.⁷⁸ Potassium penicillin G appears to have a modest, positive effect on large intestinal motility and frequency of defecation in horses.⁷⁹

MOTILITY MODIFICATION. Motility-modifying drugs could play an important role in the prevention and treatment of gastrointestinal ileus. An effective prokinetic agent could shorten the length of hospitalization, thereby reducing the cost of treatment and the number of potential complications such as weight loss, thrombophlebitis, and laminitis. There is also evidence that the development of abdominal adhesions could be minimized with the use of an effective prokinetic drug.^80,81 A potential negative effect of prokinetic use in the postoperative period after anastomosis is increased predisposi­tion to dehiscence.⁸²

For most motility-modifying drugs to be effective, the intestinal wall must be healthy in order to enhance intestinal contraction. There are several examples in horses in which the expected contractile response to certain drugs is blunted in the presence of disease. It is reasonable to assume that many putative prokinetic drugs would be partially or totally ineffective in horses with abdominal disease. Such disease would include intestine that has undergone prolonged or excessive distention or that is in any inflammatory condition, such as after intestinal manipulation or in association with primary inflammatory diseases such as DPJ or Salmonella enteritis. Numerous drugs have been investigated in humans with POI, but no prokinetic agent has yet been found to be safe and effective.⁸³ The financial pressures associated with equine patient management often compel attending veterinarians to experiment with a range of putative prokinetic drugs.

Lidocaine, an aminoamide local anesthetic, appears to be the most commonly used putative prokinetic agent in the management and prevention of POI.⁸⁴ Intravenous administra­tion of lidocaine increases gastrointestinal transit time in normal horses.⁸⁵ However, evidence regarding the efficacy of lidocaine in treating or preventing POI in horses is limited and conflicting (see below). Intravenous infusion of lidocaine may suppress primary afferent neurons, thereby limiting reflex efferent inhibition of motility. It is also possible that lidocaine may block the inhibitory effect of nonadrenergic, noncholinergic neurotransmitters on smooth muscle.⁸⁶ Lidocaine also has significant antiinflammatory properties, including amelioration of the cytokine response to endotoxemia, reduced neutrophil free radical production, impaired leukocyte phagocytic function, and inhibition of leukocyte migration through suppression of chemokines.^87,88 In addition, lidocaine has been shown to decrease clinical scores after experimental endotoxin administra­tion in healthy horses, in comparison with untreated controls.⁸⁸ Lidocaine administration can also ameliorate the inhibitory effects of flunixin meglumine on jejunal mucosal barrier recovery after ischemia.^89,90 Because systemic administration of lidocaine was shown to increase somatic thermal antinociceptive thresh- hold, but not visceral distention threshold, the beneficial effect of lidocaine in horses with gastrointestinal disease may not be associated with analgesia.⁹¹

A combination of intraoperative and postoperative lidocaine infusion is commonly used as a preventive strategy against POI on the basis of evidence suggesting that treatment with lidocaine during colic surgery improves postoperative prokinetic activity in ischemic intestine⁹² and is associated with a decreased risk of POI.³⁸ Postanesthetic treatment typically involves a slow intravenous bolus of 2% lidocaine (1.3 mg/kg) followed by a constant infusion at 0.05 mg/kg/min for 24 hours, with a target serum concentration between 1 and 2 mg/dL. In a prospective study of horses undergoing laparotomy for colic, the authors failed to demonstrate any difference between lidocaine and saline infusion with regard to return of borbo- rygmi, time to first feces, or gastric reflux.⁹³ The study was limited by a low incidence of small intestinal strangulating lesions and POI. A large prospective study in the United Kingdom did not demonstrate an effect of lidocaine treatment on the prevalence, volume, or duration of postoperative gastric reflux.⁹⁴ Lidocaine infusion was associated with a reduction in reflux duration, time to fecal passage, and hospital stay for survivors in a study of 32 horses with POI or DPJ.⁸⁶ In a larger retrospective study of 126 horses undergoing surgery for small intestinal colic, lidocaine administration was associated with a reduced incidence of POI and increased rate of short-term survival.⁴³

The rate of lidocaine infusion requires close monitoring because infusion can be associated with reversible side effects that include muscle fasciculations, ataxia, and seizures. The drug is highly protein bound, so hypoproteinemic horses may be more susceptible to toxic effects.

Bethanechol is a methyl derivative of carbachol and is an acetylcholine receptor agonist. The drug acts both at the level of the myenteric plexus and directly on intestinal smooth cells through muscarinic receptors. The actions are primarily medi­ated through activation of M₃ receptors; M₂ receptors are also involved but to a lesser extent.⁹⁵ There is evidence in other species that activation of M₂ receptors not only facilitates intestinal contraction but also may lead to antinociception.⁹⁶ Bethanechol is not degraded by the enzyme anticholinesterase. As anticipated, the drug has a range of cholinergic side effects, including abdominal discomfort, sweating, and salivation, although these are minimal when the drug is administered subcutaneously at 0.025 to 0.05 mg/kg of body weight.

Bethanechol is one of the most useful prokinetic agents in equine practice, inasmuch as it exerts effects throughout the gastrointestinal tract; however, it does not appear to be com­monly used for that purpose.⁸⁴ It is most commonly used in the management of delayed gastric emptying and slowed small intestinal transit. The drug has been shown experimentally to significantly increase gastric contractility and hasten the empty­ing of liquid- and solid-phase markers from the stomachs of normal horses.⁹⁷ The drug has potent effects in the hindgut as well. Bethanechol increases both the relative strength and duration of wall contractions in the cecum and right ventral portion of the colon and consequently speeds up cecal empty­ing.⁴⁸ When given to ponies at a dose of 0.05 mg/kg SC, the drug increased electrical activity in the large colon for approxi­mately 80 minutes.⁷⁵ Reported rates of bethanechol dosing have ranged widely; the doses most commonly recommended range from 0.025 to 0.05 mg/kg SC.

Neostigmine increases receptor levels of acetylcholine by inhibiting cholinesterase. The drug (0.022 to 0.025 mg/kg IV) promotes cecal and colonic contractile activity and enhances cecal emptying in normal ponies.⁴⁸ Neostigmine has been used in the management of small intestinal ileus but significantly delayed the emptying of 6-mm beads from the stomach of normal adult horses.⁵⁷ A survey of prokinetic use indicated that neostigmine was more commonly used in the management of large intestinal disease.⁸⁴

Metoclopramide is a moderate partial 5-hydroxytryptamine 4 (5-HT₄) receptor agonist, a moderate 5-hydroxytryptamine 3 (5-HT₃) receptor antagonist, and an antagonist of both dopamine 1 and dopamine 2 receptors. It has been suggested that the 5-HT₄ receptor agonist properties are primarily responsible for the prokinetic effects of metoclopramide. Antagonism of prejunctional dopamine 2 receptors facilitates acetylcholine release and smooth muscle contraction. The effect on dopamine receptor antagonism is absent in newer benza­mides. Metoclopramide crosses the blood-brain barrier, where its antagonist properties on central dopamine 2 receptors can result in extrapyramidal signs, including seizure. These signs were observed when the original reported dose of 0.25 mg/ kg, given by intravenous infusion over 30 minutes, was used in practice,⁹⁸ and they were responsible for poor acceptance of the drug in equine practice. An experimental bolus of endotoxin caused a delay in gastric emptying, as assessed in the acetaminophen absorption test; this delay was partially ameliorated through pretreatment with metoclopramide (0.125 mg/kg in 1 L infused over 15 minutes), but, again, the investigators reported adverse side effects.⁹⁹ Most investigators have failed to demonstrate significant effects of metoclopramide in experimental animals, but constant intravenous infusion (0.04 mg/kg/h) in a population of horses after surgery signifi­cantly decreased the volume and duration of gastric reflux in comparison to control horses and horses receiving intermittent drug infusion.¹⁰⁰ Infusion was well tolerated and was superior to intermittent infusion and to no treatment at all.

Cisapride is a second-generation benzamide that is commonly reported to act as a primary 5-HT₄ receptor agonist and 5-HT₃ receptor antagonist. Furthermore, stimulation of 5-HT₄ receptors within the enteric nervous system enhances release of acetyl­choline from the myenteric plexus, thereby promoting intestinal contraction. In vitro studies of equine jejunum concluded that the contractile actions of cisapride in the species were mediated primarily through 5-HT₂ and not 5-HT₄ receptors and that the response was noncholinergic.¹⁰1 Several reports suggested efficacy of cisapride in the management of intestinal disease in horses, including the resolution of persistent large colon impaction, treat­ment of equine grass sickness, and prevention of POI in horses after small intestinal surgery.^102-105 Cisapride has the potential to cause adverse cardiac effects, mediated through blockage of the rapid component of the delayed rectifier potassium current; those effects include lengthening of the QT interval and development of torsades de pointes, a potentially fatal arrhythmia. As a result, the drug has been withdrawn in the United States and in many other countries.

Another benzamide, mosapride, resulted in increased myoelectric activity of the small intestine and cecum of horses after oral administration. The authors reported that mosapride is a selective 5-HT₄ receptor agonist; consequently, the reported effects are surprising given that the actions of 5-HT on the equine jejunum were reported to be primarily mediated through activation of 5-HT₂ and 5-HT₃ receptors.¹⁰¹ Further work has demonstrated increased gastric emptying, jejunal and cecal motility after mosapride administration (2 mg/kg PO) to healthy horses.¹⁰⁶ Mosapride (1.5 mg/kg PO) was also shown to increase electrointestinographic maximum amplitude within the small intestine, in comparison to untreated controls, after jejunoce- costomy in normal horses.¹⁰⁷

Tegaserod, a selective 5-HT₄ receptor agonist, is used in humans for the treatment of irritable bowel syndrome.^104,108 In horses, tegaserod has been shown to decrease total gastrointestinal transit time¹⁰⁹ and increase pelvic flexure contractile activity in vitro.^110,111 Tegaserod appears safe when administered orally to horses, and plasma concentrations considered therapeutic, according to in vitro work, were obtained after administration of a single dose orally at 0.27 mg/kg.¹¹⁰

Domperidone acts as a competitive antagonist at peripheral dopamine 2 receptors; these receptors are inhibitory. The drug has been used to manage gastroparesis in humans for many years.¹¹² The primary use in equine practice is in the management of mares grazing endophyte-infected tall fescue (1.1 mg/kg/day PO), principally because of drug-enhanced prolactin release from the anterior pituitary gland. The potential prokinetic effects of domperidone have not been studied extensively in horses, but modest efficacy of domperidone (0.2 mg/kg IV) was demonstrated in a model of experimental ileus in two ponies.¹¹³

Erythromycin is a direct motilin receptor agonist acting directly on smooth muscle cells, as well as within the enteric nervous system, to facilitate the release of acetylcholine and motilin. Erythromycin enhances gastric emptying in normal horses but, in contrast to most other species, has a more pronounced effect on the hindgut.^97,114 This is somewhat surpris­ing because a higher density of motilin receptors was reported in the duodenum than in either the cecum or pelvic flexure of adult horses.¹¹⁵ Erythromycin lactobionate (1 mg/kg IV) hastens cecal emptying in normal animals and induces propagat­ing colonic MMC-like activity across the colon. Administration is often followed by defecation and abdominal discomfort. Several problems limit the use of erythromycin as a prokinetic agent in equine practice. Its principal use in human medicine has been restricted to management of acute exacerbations of diabetic gastroparesis, facilitation of feeding tube placement, and upper gastrointestinal endoscopy.¹¹⁶ Although it has been shown to enhance emptying in normal horses, it was not as effective as bethanechol.⁹⁷ Because of its potent prokinetic effects in the cecum, erythromycin may help prevent cecal impaction in horses after anesthesia. However, its effectiveness on cecal motility appears to be markedly reduced in the immediate postoperative period.³² High doses, constant infusion, or prolonged use of erythromycin also induces receptor tachyphylaxis, and its efficacy is thereby reduced. There is also evidence that intraluminal distention for as little as 2 hours leads to a reduction in the total number of motilin receptors and in the amount of motilin receptor messenger RNA, although erythromycin binding to remaining motilin receptors is not affected.¹¹⁷ Erythromycin can induce diarrhea in adult horses; therefore administration over many days should be avoided. The diarrhea induced by erythromycin is probably mediated by C. difficile overgrowth and toxin elaboration.¹¹⁸

Because opiates have an inhibitory effect on normal intestinal motility, opioid antagonists may have the potential for prokinetic activity. Naloxone (0.05 mg/kg IV) has been shown to induce contractile activity in the cecum and the left portion of the colon.¹¹⁹ The administration of naloxone was often followed by defecation within 15 to 20 minutes. Naloxone has not been beneficial in preventing POI in human beings.¹²⁰ Peripherally acting opioid antagonists, such as V-methylnaltr exone, can improve gastrointestinal tract function in humans who require opioids for pain management. An in vitro study in horses demonstrated a direct contractile effect of V-methylnaltrexone on the circular muscle layer of the jejunum and large colon.¹²¹ In healthy horses, V-methylnaltrexone therapy increased daily fecal weight and, when administered concurrently with mor­phine, partially prevented morphine-associated increases in intestinal transit time and decreases in fecal production.¹²²

K₂-Adrenoreceptor antagonists, such as yohimbine and tolazoline, counteract increased sympathetic outflow in response to nociceptive stimulation. Yohimbine (75 μg7ltissues as the gastrointestinal mucosa and kidneys and in the endothelium and platelets. In contrast, COX-2 is primarily an inducible enzyme that has a critical role in inflam­mation and is produced by a variety of cells, including monocytes, neutrophils, epithelial cells, fibroblasts, synoviocytes, and chondrocytes.

It has been postulated that drugs that nonselectively inhibit both COX-1 and COX-2 have greater toxic potential because they inhibit prostaglandins necessary for physiologic homeo­stasis, as well as prostaglandins that mediate inflammation and pain.⁴ Both COX-1- and COX-2-dependent prostaglandins play an important role not only in maintaining the epithelial barrier in the gut but also in healing the epithelium when the mucosa is damaged (e.g., because of ischemic injury or infection). Most of the commonly used NSAIDs in horses are considered to be nonselective. Exceptions include meloxicam (3.8 to 10 times more selective for equine COX-2 versus COX-1^5,6) and firocoxib (268 to 643 times more selective for COX-2 than for COX-1⁷). NSAIDs that are COX-2 selective are less ulcerogenic than those that are COX-1 selective in other species and may be so in horses. For example, COX-2 selectivity decreases the adverse effects of NSAIDs in a model of epithelial repair in horses.⁸

Although the COX-1 versus COX-2 scheme is currently considered valid, some evidence indicates that it may be overly simplistic. For example, COX-1 may play an important role in inflammation and is at least partly inducible.⁹ In contrast, COX-2 can be induced physiologically in various organs and tissues and by stimuli other than inflammation.^4,10 In horses, gastric ulcerogenicity of even the nonselective NSAIDs varies (phenylbutazone is more ulcerogenic than flunixin meglumine, which is more ulcerogenic than ketoprofen).¹¹ This difference in toxicity among drugs may relate not only to the COX selectivity but also to other factors such as tissue distribution.

The gastrointestinal tract and the kidneys are the most common targets for NSAID toxicity. NSAID-induced injury can develop anywhere in the gastrointestinal tract (from the mouth to the rectum). Two well-recognized syndromes may be attributed, at least in part, to NSAID toxicity. The first is gastric ulceration: In the stomach, inhibition of COX can increase acid secretion, decrease output of mucus and bicarbon­ate, impair vasodilation, and diminish epithelial restitution, cell division, and angiogenesis.¹² Inhibition of COX also impairs the healing of existing ulcers. The second syndrome attributed to NSAID toxicity is right dorsal ulcerative colitis (RDUC). Although the right dorsal segment of the large intestine is most commonly affected, other segments may also be involved. Ulcerative lesions in the large intestine can be particularly troublesome because they can cause chronic debilitation, are difficult to diagnose, and can be refractory to treatment. In the kidney, prostaglandin E₂ and prostacyclin (prostaglandin I₂) produce vasodilation in the autoregulatory response of renal blood flow to hypoperfusion; consequently, hypovolemia, hemorrhage, or renal disease increases the risk of renal NSAID toxicosis. Damage is greatest at the renal crest (papilla), and papillary crest necrosis may be associated with subsequent nephrolithiasis or ureterolithiasis and chronic renal failure.¹³ In humans the most common side effect of NSAIDs is bleeding, caused in part by reduced function of platelets and in part by gastrointestinal hemorrhage.

Not all of the adverse effects of NSAIDs are attributable to COX inhibition. The NSAIDs also cause injury through a variety of mechanisms, including microvascular damage, increased intracellular concentration of reactive oxygen and other free radicals, direct local injury (particularly with ion trapping in the stomach), inhibition of cell division, and reduced hydro­phobicity of the gastric mucous coat.^10,12 Inhibition of COX may cause arachidonic acid metabolism to be shunted toward the lipoxygenase pathway, thereby producing other biologically active eicosanoids. The clinical significance of this shunting is unclear, but it has the potential for deleterious effects.

Although the toxicity of NSAIDs is related to the dose and duration of administration, some horses develop toxicosis at recommended doses. Predisposing factors such as dehydration, renal disease, hepatic disease, or sepsis may contribute to the development of NSAID toxicity. Dehydration, renal disease, and hepatic disease predispose horses to NSAID toxicosis because of reduced tissue perfusion and reduced drug elimina­tion. Sepsis may predispose horses to NSAID toxicosis because of secondary hypovolemia, decreased tissue perfusion, and direct and indirect effects of various mediators produced in response to sepsis (e.g., platelet-aggregating factor). In humans, risk of NSAID-induced ulceration is increased among those with various gastrointestinal disorders (e.g., IBDs). NSAIDs inhibit the ability of injured equine intestinal mucosa to repair, which may increase the risk of ulceration in horses with ischemic damage or intestinal infections. Body weight may be a predispos­ing factor in that NSAIDs are often administered to ponies, miniature horses, and small horses at doses higher than those recommended for their body weight. Inadvertent overdosing can occur regardless of body weight or size (e.g., administration of a 12-g tube of phenylbutazone paste when administration of an anthelmintic paste was intended).

Some horses may have an idiosyncratic predisposition for ulceration, particularly that of the right dorsal portion of the colon. In experimental conditions, arthritic laboratory animals were more susceptible to NSAID-induced gastropathy than were healthy animals.¹⁴ This finding may have relevance to horses because NSAIDs are often administered to chronically lame horses. Two or more NSAIDs are used concurrently in some situations. It is important to recognize that the effects of combining NSAIDs are additive, so that administering two NSAIDs at each of their recommended doses is similar to giving twice the recommended dose of one NSAID. Combina­tion of two NSAIDs prolongs their pharmacologic effect but increases the risk of toxicity.¹⁵

■ ClinicalSigns Clinical signs of NSAID toxicity are usually referable to the alimentary system and vary, depending on the segment involved. Oral or lingual ulceration may lead to difficulty in prehension and mastication. Esophageal ulceration may result in excessive salivation and apparent signs of pain (stretching of the neck, groaning) during swallowing. Gastric ulceration may result in slow consumption of feed, inappetence (particularly for grain in some horses), or anorexia. Horses that have gastric outflow obstruction associated with gastroduodenal ulceration may exhibit ptyalism, reflux esophagitis, and, in severe cases, spontaneous nasogastric reflux. Horses with ulceration anywhere in the gastrointestinal tract may exhibit signs of colic, which may be intermittent and may vary in severity. Horses with colonic ulceration may have soft stool or diarrhea and ventral edema secondary to enteric protein loss. Diarrhea can be severe, even fatal. Sepsis may result from intestinal mucosal damage caused by NSAIDs. Clinical signs of sepsis (e.g., tachycardia, altered appearance of mucous membranes, fever, and dehydration) may be seen in some horses with NSAID enteropathy. In some horses, hematuria may be seen.

Horses may have clinical signs days to weeks after NSAID therapy. Such horses are typically brought to medical attention because of recurring colic, weight loss, or loose manure. It is particularly important in these horses to determine any history of NSAID administration, even if it was several weeks before the time of presentation.

■ Diagnosis Diagnosis is usually made on the basis of history of NSAID use, clinical signs, and clinicopathologic findings. The most consistent clinicopathologic abnormalities in horses with NSAID toxicosis are hypoproteinemia and hypoalbuminemia, presumably resulting from loss of protein through inflamed or ulcerated intestinal mucosa. These findings are more commonly observed with damage to the distal intestinal tract and are not reliable for diagnosis of NSAID gastropathy. Some horses have decreased serum concentration of calcium, presumably attributable to intestinal loss of protein-bound calcium. In horses with severe NSAID-induced diarrhea, hyponatremia, hypochloremia, hypokalemia, acidemia, and hypovolemia may be observed. In such cases, hypovolemia may make the serum protein concentration appear to be higher than its actual value would be if the horse were adequately hydrated.

In chronic cases, horses may be anemic from inflammation, intestinal loss of blood through ulceration, or reduced function of platelets. Occult blood may be found in the feces of horses with lesions in the more distal portions of the intestinal tract. Tests for occult blood often lack sensitivity, and false-positive results may be expected for up to 24 hours after rectal palpation.

The concentration of leukocytes is usually within the refer­ence range, although leukocytosis and hyperfibrinogenemia, associated with inflammation, and leukopenia and neutropenia, presumably caused by sepsis, can occur in some horses with acute NSAid toxicosis that affects the distal intestine. In general, results of peritoneal fluid analysis are within reference ranges, but concentration of nucleated WBCs, total protein, and fibrinogen may increase in the presence of advanced intestinal damage or intestinal vascular infarction. When findings of cytologic examination of peritoneal fluid are abnormal, results are more consistent with nonseptic than septic inflammation; however, septic inflammation may be observed when severe intestinal ulceration leads to transmural lesions and septic peritonitis.

Several clinicopathologic changes may accompany NSAID- induced renal damage. The most consistent finding is a decrease in urine specific gravity, from 1.008 to 1.020. Inability to properly dilute urine can occur with acute NSAID toxicosis for years after the original insult. This results from preferential damage to areas of the kidney that contribute most to con­centrating urine (medulla, papilla). In chronic cases, urine specific gravity typically ranges from 1.013 to 1.020. Some horses with NSAID toxicosis are azotemic. In acute cases azotemia can result from dehydration, NSAID-induced altera­tions in renal blood flow, and tubuloglomerular feedback mechanisms. Chronic azotemia, in which serum creatinine levels range from 2.1 to 3.5 mg/dL, results from tubuloglo- merular feedback mechanisms that reduce glomerular filtration to compensate for reduced reclamation of solutes in the medullary collecting ducts. In acute NSAID toxicosis, affected animals may have overt hematuria. In other cases, urinalysis may reveal occult blood, increased renal cells, and increased WBCs. In chronic cases other than those involving decreased urine specific gravity, urinalysis results are typically normal.

Endoscopy can be useful for visualizing the location and extent of esophageal and gastric lesions. NSAID-induced gastric lesions are more common in the glandular epithelium, although nonglandular lesions can be observed. Contrast radiography or scintigraphy may be useful for documenting delayed gastric emptying in some horses. Lesions of the jejunum, ileum, cecum, and colon can be difficult to identify without celiotomy and enterotomy. Isotope-labeled WBC scintigraphic scans may identify colonic ulceration¹⁶; the sensitivity and availability of the procedure, however, are limited. Ultrasonography may reveal thickening of the right dorsal portion of the colon or other colonic segments, but the technique appears to lack sensitivity.¹⁷ Horses with renal crest necrosis may exhibit increased ultrasonographic echogenicity of the renal crest and echogenic debris in the renal pelvis.

■ Management Administration of NSAIDs should be discontinued if NSAID toxicosis is suspected. Gastric lavage and administration of 1 gallon of mineral oil per 450 kg of body weight via nasogastric tube may be of benefit in horses with acute NSAID overdose to reduce the absorption of the administered NSAID. Treatment for gastric ulceration with a proton-pump inhibitor (e.g., omeprazole, an H2-receptor blocker such as ranitidine, or sucralfate, or a combination of these) should be implemented in horses with gastric ulceration. Sucralfate (10 to 20 mg/kg PO q8-12h) may also have a benefit in more distal segments of the equine intestine affected by NSAID toxicity. Sucralfate reduced ulceration in the stomach, small intestine, and colon and reduced protein loss in weanling­aged horses treated with an overdose of phenylbutazone.¹⁸

Regardless of the site of NSAID toxicity, administra­tion of prostaglandin E analogs, or drugs that mimic their effects, may be of benefit. Administration of a synthetic analog of prostaglandin E₂ has been demonstrated to prevent phenylbutazone-induced gastrointestinal lesions in horses.¹⁹ Misoprostol, a synthetic analog of prostaglandin E₁, ameliorated the effect of flunixin and promoted in vitro repair of equine intestinal mucosa damaged by ischemia.²⁰ The beneficial effects of misoprostol are probably due in part to the beneficial effects of prostaglandin E on the gastrointestinal mucosa (reducing injury and increasing repair) but may also be attributed to the antiinflammatory properties of misoprostol, particularly in the stomach, where misoprostol concentrations are high after oral administration.^21,22 Misoprostol can be administered orally starting at doses of 5 ug/lintestinal fill in the colon. A diet lower in fiber should decrease the physiologic load of the colon because in horses the cecum and large colon are the primary sites of fiber digestion and exchange of fluid and electrolytes. Concentrate should be fed in smaller amounts and frequently (four to six feedings per day). Addition of corn oil may provide additional calories and may also aid in healing of the damaged intestinal mucosa by promoting prostaglandin E₂ production. Some horses will not eat complete pellets, and some horses that have roughage withheld will eat bedding or wood as a consequence. Such horses should be allowed to eat fresh grass in small amounts on a frequent basis (four to six times daily). The importance and optimal duration of restriction of roughage are unknown, but it probably requires months for the colon to heal. Horses should be changed from, and returned to, their usual diet over a period of several days to decrease the risk of inducing other digestive disorders.

Feeding psyllium mucilloid may promote colonic healing in horses with RDUC. In other animal species, psyllium mucilloid has been demonstrated to increase the concentration of short-chain fatty acids of the large bowel, and increased short-chain fatty acids can promote colonic mucosal repair.²⁴ The amount and duration of psyllium mucilloid administered orally that are necessary to alter the colonic concentration of short-chain fatty acids and the role of short-chain fatty acids in repair of RDUC in horses are unknown. Continuous feeding according to manufacturer's recommendations for 3 to 6 months is suggested, or feeding 1 to 2 oz of psyllium mucilloid once or twice daily for the same duration may be considered.

Horses with strictures of the pylorus, duodenum, jejunum, or colon may require surgical management. Bypass or resection of affected intestinal segments may be necessary.

Limiting the extent of predisposing factors, such as dehydra­tion, should decrease the risk of NSAID toxicosis. Avoiding use of NSAIDs or limiting the dose and duration of treatment to the minimums necessary to control the primary problem is recommended to decrease the risk of NSAID toxicosis. Other approaches to analgesia, such as regional anesthesia (epidural or perineural nerve blocks of distal limbs) or administration of butorphanol or lidocaine via constant-rate infusion should be considered.