The toxicity of NSAIDs is related to COX selectively, dosage, and duration (see Nonsteroidal Anti-inflammatory Drugs). It is hypothesized that nonselective COX inhibitors have greater risk of toxicity than COX-selective drugs. COX inhibitors also cause delayed GI healing. The GI tract and kidneys are most commonly affected by NSAID toxicity. NSAID-induced injury can occur anywhere in the GI tract, but the large colon (especially the right dorsal colon) and gastric mucosa appear to be the most sensitive. The ulcerogenicity of phenylbutazone is greater than that of flunixin meglumine, which has greater ulcerogenicity than ketoprofen. The COX-2 selective inhibitor firocoxib appears to be safer. Ulcerative lesions in the large colon lead to a protein-losing enteropathy, often with clinical signs of ventral edema, anorexia, lethargy, weight loss, diarrhea, and colic. Scarring of the right dorsal colon can occur leading to large-colon impactions, sometimes requiring large-colon resection.
Phenylbutazone administered at high doses or for prolonged periods causes a protein-losing enterocolopathy in horses. However, some horses are inherently sensitive to NSAIDs, and right dorsal colitis can occur at lower than recommended dosages. Toxicosis can develop from oral or parenteral administration of NSAIDs. Hypoproteinemia (hypoalbuminemia and hypoglobulinemia) is seen due to loss of protein into the intestinal lumen, which can occur without visible ulceration. Renal papillary necrosis may also be seen. Administration of flunixin meglumine at high doses or for prolonged periods can result in a similar toxicosis.
Clinical signs of NSAID toxicity include difficulty in mastication due to oral and lingual ulceration, hypersalivation, and signs of pain when swallowing due to esophageal ulceration. Gastric ulceration can result in recumbency after eating, signs of colic, and anorexia. Horses with colonic ulceration can have soft feces, diarrhea, and ventral edema. Intestinal ulceration can be severe enough to allow endotoxin and bacterial translocation and signs of systemic inflammation and septicemia. Dehydration, fever, and tachycardia can occur in severe cases. Clinical signs can occur days to weeks after NSAID therapy. More chronic cases present with recurring colic, weight loss, and soft feces.
A tentative diagnosis can be made based on history of NSAID administration, clinical signs, and presence of hypoproteinemia. Severe cases may have hyponatremia, hypochloremia, hypocalcemia, and acidemia in addition to hypovolemia. Ultrasonography may detect thickening of the colon. Gastric ulceration can be confirmed by gastroscopy but requires an endoscope 2–3 m long.
Treatment includes discontinuing use of phenylbutazone or any other NSAID. In acute toxicosis, administration of 1 gal. of mineral oil repeated after 2 hr may be beneficial to decrease drug absorption. To help prevent gastric ulceration, reducing production of gastric acid with an H2-receptor blocker (eg, ranitidine) or a proton pump inhibitor (eg, omeprazole) may be beneficial; sucralfate may be indicated as well. Administration of misoprostol (a synthetic prostaglandin analogue) may be beneficial but can cause additional signs of diarrhea and colic. IV fluid therapy is indicated in cases of hypovolemia, especially with concurrent azotemia. Plasma transfusion or synthetic colloids can be used to increase plasma oncotic pressure.
Longterm dietary management consisting of a low-fiber complete pelleted ration fed several times throughout the day, and elimination of roughage is recommended. Corn oil can be given to provide supplemental calories and may help heal damaged intestinal mucosa. Psyllium mucilloid can also promote colonic healing by increasing the concentration of short-chain fatty acids. Surgery may be required if scarring of the bowel has resulted in partial obstruction of the intestine.
Prevention of NSAID toxicity involves limiting the dose and duration of NSAID treatment, using a COX-2 selective NSAID, or relying on alternative analgesic therapy. Monitoring for fecal consistency and serum albumin concentrations are easy methods to detect the development of right dorsal colitis in horses receiving NSAIDs.