Abstract
OBJECTIVE
To report the incidence and characteristics of gastrointestinal ulceration lesions in dogs receiving an NSAID and/or corticosteroid.
ANIMALS
33 dogs.
CLINICAL PResentation
Medical records of dogs with gastrointestinal ulceration receiving NSAIDs and/or corticosteroids within 30 days of diagnosis between January 2012 and July 2022 at multiple referral institutions were reviewed. Diagnosis was confirmed via endoscopy, surgery, or necropsy. Clinical data were collected from the medical record, including the dose and reason for administration of NSAIDs or steroids.
RESULTS
Dogs received a single NSAID (n = 22, most commonly carprofen [9], meloxicam [4], and deracoxib [3]), 2 NSAIDs (5), a single steroid (5: prednisolone [2], prednisone [2], or dexamethasone SP [1]), or an NSAID and steroid (1). Eleven dogs receiving a single cyclooxygenase (COX)-2–sparing NSAID at an appropriate dose had ulcerations. All dogs receiving 2 NSAIDs concurrently experienced full-thickness perforation (5 of 5). The most common ulcer locations were duodenum (n = 18) and pylorus (11). Abdominal ultrasound correctly identified the site of ulceration in 5 of 24 dogs.
CLINICAL RELEVANCE
Dogs receiving COX-2 sparing NSAIDs at recommended doses are at risk of severe GI ulceration. Carprofen was the most common NSAID resulting in ulceration; however, it is one of the most prescribed NSAIDs. Adding another NSAID and steroid could increase this risk. Careful monitoring is crucial for dogs on NSAIDs, regardless of duration.
Introduction
Gastrointestinal (GI) ulceration in dogs is a significant complication that can be associated with systemic disease such as neoplasia, inflammatory bowel disease, hepatic dysfunction, hyperacidity, pancreatitis, or hypoadrenocorticism and various drug therapies.1–3 Nonsteroidal anti-inflammatory drugs and corticosteroids are mainstay therapies for multiple diseases in veterinary medicine, despite their link to GI ulceration. Nonsteroidal anti-inflammatory drug administration is one of the most common causes of GI ulceration in dogs.4 Nonselective and selective cyclooxygenase (COX) NSAIDs exist, with the former posing the greatest risk of ulceration and thus warranting cautious use.1 Nonsteroidal anti-inflammatory drugs lead to GI ulceration through direct topical effect and the inhibition of the COX-1 and COX-2 pathways.5–7 The COX pathways allow for the formation of prostaglandins that are essential for gastroprotection. Inhibition of prostaglandin production decreases mucosal blood flow, mucus production, bicarbonate secretion, and epithelial turnover.8 In contrast, the mechanism of GI ulceration due to steroid administration remains poorly understood. It has been suggested that ulceration is due to the inhibition of phospholipase A, which causes reduction in cytoprotective prostaglandins that repair the gastric mucosa.9
In recent years, COX-2–selective NSAIDS have been promoted due to their selectivity and reduced risk of GI ulceration; however, they still carry the risk of GI ulceration even when administered at appropriate doses.4 Gastrointestinal ulcerations were observed via endoscopy in 10 out of 12 dogs undergoing chronic treatment at an appropriate dose with COX-2–selective NSAIDs in 1 study.4 Similarly, healthy dogs receiving corticosteroids are 7 to 11 times as likely to have higher endoscopic lesion scores compared to control groups, when those corticosteroids are administered at a dose of 2 mg/kg/d.10 Both NSAIDs and corticosteroids can induce lesions in otherwise healthy dogs, which may persist without showing apparent clinical symptoms.10–12
The clinical signs of GI ulceration can vary widely, ranging from overt GI symptoms to weight loss, typically depending on the chronicity of the condition. Commonly reported clinical signs include abdominal pain, hematemesis, melena, and anorexia.3,6,13 Clinicopathologic findings are seldom specific to this disease, with the most common being regenerative or nonregenerative anemia or reticulocytosis without anemia, depending on the severity and duration of the disease.1 Subtle physical examination findings and nonspecific diagnostic changes can contribute to the underdiagnosis of this condition.13,14 Imaging modalities such as abdominal radiographs and ultrasound often yield variable diagnoses, with ulcerative lesions frequently remaining unconfirmed until endoscopy, surgery, or necropsy. The ability to detect these erosions often depends on the operator’s experience and equipment.
There are few studies in the literature reporting findings and outcomes associated with GI ulceration due to NSAIDs or corticosteroids, and most focus on small populations of dogs. The objectives of our study were to report the incidence, characteristics, and outcomes of GI ulcer lesions in a larger series of dogs treated with these medications. Additionally, we aimed to determine whether any NSAID or corticosteroid dosing regimen is overrepresented among dogs diagnosed with GI ulcers.
Methods
Dogs with the diagnosis of GI ulceration between January 2012 and July 2022 were identified by electronic search of medical records at Texas A&M Veterinary Medical Teaching Hospital, University of Georgia Veterinary Teaching Hospital, and Foster Hospital for Small Animals at Tufts University. Dogs were included if an NSAID or corticosteroid was administered within 30 days prior to diagnosis and the ulceration diagnosis was confirmed by direct identification via endoscopy, surgical exploration, or necropsy. Dogs where GI ulceration was suspected but not confirmed via one of these methods were excluded.
Data collected from the medical record for all dogs included signalment (breed, age, body weight, sex, reproductive status), current medications, concurrent medical issues, clinical signs, presenting complaint, and physical examination findings. The dosing regimen and reasons for the administration of NSAIDs or steroids were recorded as well as prior use of similar medications. Diagnostics recorded included chemistry, CBC, urinalysis, coagulation panel, abdominal radiographs, endoscopy, ultrasound, contrast radiographs, and CT scans. Dates of ulcer diagnosis and methods of definitive diagnosis were obtained, as well as the number of ulcers, size and depth of ulceration, and location. Medical therapies used as the sole treatment of ulceration or prior to surgical intervention were documented. In dogs that underwent surgery, the surgical procedures performed, intraoperative cultures, and complications were obtained. Postoperative complications and therapies, alongside the duration of stay in the hospital and survival, were recorded.
Statistics
Statistical analysis was performed in Prism, version 9.0.0 (GraphPad Software LLC). A Shapiro-Wilk normality test was performed on all data; mean and SD were reported for normally distributed data, and median and range were reported for nonnormally distributed data.15
Results
History
Thirty-three dogs were identified. Types of dogs included 6 Labrador Retrievers, 3 German Shepherd Dogs, 2 Border Collies, 2 Mastiffs, 2 Rottweilers, and 2 mixed-breed dogs, as well as 16 additional breeds with only 1 dog represented. Median weight was 31.5 kg (range, 5.5 to 109 kg) and age was 8.5 ± 3.2 years at the time of ulcer diagnosis. There were 18 males (16 neutered, 2 intact) and 15 females (all spayed).
Dogs were presented to the hospital displaying clinical signs for a median duration of 3 days (range, 1 to 60 days). The most common clinical signs were vomiting (n = 27), anorexia (26), lethargy (22), melena (6), and hematochezia (4). Of the 27 vomiting dogs, 6 were reported to have hematemesis. Concurrent medical issues included orthopedic disease (n = 9), soft tissue neoplastic mass (7, including apocrine gland adenocarcinoma [2], nasal carcinoma, a grade 2 cutaneous mast cell tumor, cutaneous soft tissue sarcoma, pheochromocytoma, and a lipoma), epilepsy (2), immune-mediated hemolytic anemia (2), diabetes mellitus with ketoacidosis (1), chronic nonregenerative anemia (1), chronic hypoalbuminemia (1), suspected tick-borne thrombocytopenia (1), chronic kidney disease (1), protein-losing enteropathy (1), proteinuria (1), exocrine pancreatic insufficiency (1), idiopathic pericardial effusion (1), nonambulatory paraparesis (1), Lyme disease (1), pneumonia (1), hypertension (1), and idiopathic pleural effusion (1). Concurrent medical issues paired with NSAID or steroid administration data for each dog were recorded (Table 1).
The dosing regimen, method of diagnosis, ulcer depth, concomitant diseases, and outcome of 33 dogs administered NSAIDs, steroids, or a combination of the two that were diagnosed with clinical gastrointestinal ulceration.
| Dog | NSAID or steroid | Dose (mg/kg/d) | Route | Duration (d) | Method of ulcer diagnosis/depth | Concurrent diseases | Outcome |
|---|---|---|---|---|---|---|---|
| 1 | Deracoxib | 1.5 | PO | 270 | Surgery/Perf | OA, pneumonia, idiopathic epilepsy | Died |
| 2 | Firocoxib | 6.9 | PO | 21 | Surgery/Perf | Chronic hypoalbuminemia, OA | Survived |
| 3 | Prednisolone | 1.6 | PO | 120 | Endoscopy/Supf | IMHA, severe proteinuria | Survived |
| 4 | Firocoxib | 4.4 | PO | 21 | Surgery /Perf | Nasal carcinoma | Survived |
| 5 | Piroxicam | 0.3 | PO | 73 | Surgery/Perf | AGASACA | Survived |
| 6 | Carprofen | 3.5 | PO | 17 | Necropsy/Perf | Pheochromocytoma | Died |
| 7 | Carprofen | 2.6 | PO | 73 | Surgery/Perf | Soft tissue sarcoma | Died |
| 8 | Meloxicam | 0.18 | PO | 21 | Necropsy/Perf | OA | Died |
| 9 | Prednisolone | 2 | PO | 1 | Necropsy/Supf | Lipoma, thrombocytopenia | Died |
| 10 | Carprofen | 4.2 | PO, SC | 6 | Necropsy/Supf | Osteosarcoma (NSAID after limb amputation), chronic kidney disease, systemic hypertension | Died |
| 11 | Meloxicam, Carprofen | 0.1, 4.8 | SC, PO | 6, 47 | Necropsy/Perf | Mast cell tumors on elbow and scapula | Died |
| 12 | Meloxicam | 0.17 | PO, SC | 6 | Surgery/Perf | — | Survived |
| 13 | Meloxicam, ibuprofen | Unknown, 65 | PO | 5, 5 | Surgery/Perf | Nonambulatory paraparesis | Died |
| 14 | Meloxicam | 0.1 | PO | 152 | Surgery/Perf | OA | Survived |
| 15 | Piroxicam | 5 | PO | 15 | Surgery/Perf | — | Died |
| 16 | Carprofen | 3.5 | PO | 6 | Surgery/Perf | AGASACA | Survived |
| 17 | Deracoxib | 0.94 | PO | 14 | Necropsy/Perf | Obese | Died |
| 18 | Dex. SP | 0.1 | IV | 2 | Surgery/Perf | Cholecystectomy 3 wk prior, bacteriuria | Died |
| 19 | Carprofen | 2 | PO | 4 | Surgery/Perf | Idiopathic pericardial effusion (pericardial window 5 d prior) | Survived |
| 20 | Deracoxib, Carprofen | 2, 4.1 | PO, SC | 7, 1 | Surgery/Perf | — | Died |
| 21 | Deracoxib, Carprofen | 1.79, 3.6 | PO | 59, 2 | Surgery/Perf | Osteosarcoma | Died |
| 22 | Ketoprofen | 5.2 | PO | 6 | Surgery/Perf | — | Survived |
| 23 | Carprofen | 5.3 | PO | 14 | Surgery/Perf | Idiopathic pleural effusion | Survived |
| 24 | Deracoxib | 0.92 | PO | 8 | Surgery/Perf | Lyme positive, OA | Survived |
| 25 | Prednisone | 1.8 | PO | 11 | Surgery/Supf | IMHA, prior gastrotomy with jejunal intussusception and second surgery 2 d later | Survived |
| 26 | Carprofen, Prednisone | Unknown, 0.9 | PO | Unknown, 7 | Necropsy/Supf | Exocrine pancreatic insufficiency, OA | Died |
| 27 | Carprofen | 3.97 | PO | 5 | Necropsy/Supf | — | Died |
| 28 | Prednisone | 0.52 | PO | 13 | Endoscopy/Supf | Protein-losing enteropathy, idiopathic epilepsy | Died |
| 29 | Meloxicam | Unknown | PO | 14 | Surgery/Perf | OA | Survived |
| 30 | Aspirin | 9.8 | PO | 2 | Endoscopy/Supf | OA, chronic weight loss | Survived |
| 31 | Firocoxib | Unknown | PO | > 365 | Endoscopy/Supf | Chronic vomiting for 6 wk before diagnosis | Survived |
| 32 | Carprofen | 1.7 | PO | 9 | Surgery/Supf | OA | Survived |
| 33 | Carprofen | Unknown | PO | 3 | Endoscopy/Supf | Diabetes mellitus with DKA | Survived |
— = No concurrent diseases reported. AGASACA = Apocrine gland anal sac adenocarcinoma. Dex. SP = Dexamethasone SP. DKA = Diabetic ketoacidosis. IMHA = Immune-mediated hemolytic anemia. OA = Osteoarthritis. Supf = Superficial. Perf = Perforating.
Concurrent non-NSAID or steroid medications at the time of ulcer diagnosis included gabapentin (n = 10), tramadol (9), maropitant (7), ampicillin (4), doxycycline (4), omeprazole (4), sucralfate (4), clopidogrel (4), enrofloxacin (3), cephalexin (2), levetiracetam (2), diphenhydramine (2), and a variety of other medications received by 1 dog only.
Examination and diagnostic findings
The most common physical examination abnormality was dull mentation (n = 13). Additional examination findings included abdominal pain (n = 9), abdominal distension (7), pale mucous membranes (6), dyspnea (5), and pyrexia (5). Abdominal radiographs were performed in 13 dogs, with reports available for review in 12 dogs. Decreased serosal detail was noted in 9 dogs, pneumoperitoneum was noted in 4 dogs, and radiographs were reported as normal in 3 dogs. Abdominal ultrasound was performed in 24 dogs, with identification of a site of ulceration in only 5 dogs. Peritoneal effusion was identified in 18 dogs, thickening of a portion of the GI tract in 13 dogs, and pneumoperitoneum in 8 dogs.
The most common clinicopathologic abnormalities were anemia (n = 16), azotemia (13), hyperlactatemia (13), and hypoalbuminemia (11). Select biochemistry and CBC information is available in Supplementary Table S1. Twenty-two dogs were diagnosed with septic peritonitis. Eighteen of these dogs were treated with surgery, and the remaining 4 were euthanized. Of the 22 dogs with septic peritonitis, 13 had preoperative or pre-euthanasia abdominocentesis with cytology performed. Neutrophils were the predominate cell type in all cytologies. Intracellular bacteria were definitively identified in 4 cytology samples; in 7 samples it was noted that no bacteria were seen (but a perforated ulcer was later confirmed), and in 2 samples there were possible bacteria seen. The location of perforation in cases with bacteria not definitely seen on cytology included the proximal duodenum (n = 4), pylorus (3), greater curvature of the stomach (1), and both pylori extending into the proximal duodenum (1). One sample each had possible bile material, yeast, and mycotic species visualized. Paired glucose and lactate analysis was performed in 3 dogs; median whole blood–peritoneal fluid glucose was 32 mg/dL (range, 28 to 57 mg/dL), and median whole blood–peritoneal fluid lactate was 4.6 mmol/L (range, 6.8 to 4.3 mmol/L).
NSAID or steroid history
Twenty-two dogs were receiving a single NSAID at the time of ulcer diagnosis, 5 dogs received 2 different NSAIDs within 14 days of ulcer diagnosis, 5 dogs were receiving a steroid only, and 1 dog was receiving an NSAID and a steroid concurrently. Reasons for NSAID administration included orthopedic disease (n = 7), pain control in the postoperative period (7), back pain (2), wounds (2), cervical pain (1), rhinitis (1), pain associated with an anal sac mass (1), pain associated with a soft tissue sarcoma (1), dental disease (1), pancreatitis (1), pelvic limb swelling (1), Lyme disease (1), and unknown (1). Surgeries performed in dogs receiving a postoperative NSAID included limb amputation for osteosarcoma (2 dogs), cutaneous mast cell tumor removal (1 dog), anal sacculectomy combined with liver lobectomy (1 dog), pericardial window (1 dog), tongue laceration repair (1 dog), and stifle surgery (1 dog). Reasons for steroid administration included immune-mediated hemolytic anemia (2 dogs), active hepatitis (1 dog), cervical spondylomyelopathy (1 dog), and thrombocytopenia (1 dog). The dog receiving an NSAID and steroid concurrently was on the NSAID for orthopedic disease and the steroid for unknown reasons.
Details on drug distribution and doses of NSAID and/or steroid administered are available in Table 1, Table 2, and Table 3. In dogs receiving a single NSAID only, the drug administered was carprofen (n = 9), meloxicam (4), deracoxib (3), firocoxib (2), piroxicam (2), ketoprofen (1), and aspirin (1). Five dogs receiving an NSAID at the time of ulcer diagnosis had a history of administration of another NSAID class within the 30 days prior. The combination included deracoxib with a single concurrent carprofen injection in 1 dog, deracoxib with carprofen following a 3-day washout period in 1 dog, firocoxib with a single concurrent meloxicam injection in 1 dog, and carprofen with a single concurrent meloxicam injection in 1 dog; 1 dog on meloxicam had been given ibuprofen by the owner for 5 days ending 2 weeks prior to starting meloxicam. In dogs receiving a steroid only, the drug administered was prednisolone (n = 2), prednisone (2), and dexamethasone sodium phosphate in 1 dog. The dog receiving concurrent NSAID and steroid therapy was receiving carprofen and prednisone for 8 days.
The dosing regimen, recommended dosage, and number of dogs receiving above the recommended dosage in 33 dogs receiving NSAIDs or steroids that resulted in clinical gastrointestinal ulceration.
| Drug | No. of dogs | Published dose ranges* | Dose administered (median and range) | No. of dogs receiving above recommended dose |
|---|---|---|---|---|
| NSAIDs | ||||
| Carprofen | 13* | 4.4 mg/kg/d | 3.6 mg/kg/d (range, 1.7–5.3 mg/kg/d; dose unknown in 2 dogs) | 2† |
| Meloxicam | 6* | 0.1 mg/kg/d with optional 0.2-mg/kg loading dose | 0.14 mg/kg/d (range, 0.1–0.18 mg/kg/d; dose unknown in 3 dogs) | 2‡ |
| Deracoxib | 5* | 1–2 mg/kg/d, or 3–4 mg/kg/d for max of 7 d | 1.5 mg/kg/d (range, 0.9–2.0 mg/kg/d) | 0§ |
| Firocoxib | 2* | 5 mg/kg/d | 4.4 and 6.9 mg/kg/d | 1** |
| Piroxicam | 2 | 0.3 mg/kg/d | 0.3 and 5 mg/kg/d | 1 |
| Ketoprofen | 1 | 0.25–1 mg/kg/d, or 2 mg/kg/d up to 3 d described (not recommended) | 5.2 mg/kg/d | 1 |
| Aspirin | 1 | Up to 40 mg/kg/d described (not recommended) | 9.8 mg/kg/d | 1 |
| Steroids | ||||
| Prednisone | 3* | Up to 4 mg/kg/d | 0.9 mg/kg/d (range, 0.52–1.8 mg/kg/d) | 0** |
| Prednisolone | 2 | Up to 4 mg/kg/d | 1.6 and 2.0 mg/kg/d | 0 |
| Dex. SP | 1 | Up to 1 mg/kg/d | 0.1 mg/kg/d | 0 |
Dosages are reported as median and range.
*Indicates dogs that received a concurrent NSAID or steroid; recommended dosages are based off Plumb’s Veterinary Drug Handbook.
†Concurrent deracoxib in 2 dogs, meloxicam in 1 dog, prednisone in 1 dog.
‡Concurrent firocoxib in 1 dog, concurrent carprofen in 1 dog, recent ibuprofen in 1 dog.
§Concurrent carprofen in 2 dogs.
**Concurrent meloxicam in 1 dog.
The dosing regimen, washout period or overlap, and reason for drug administration in 6 dogs that received either 2 NSAIDs in combination or an NSAID in combination with a steroid, resulting in clinical gastrointestinal ulceration.
| Dog | Drug 1 | Drug 2 | Washout or duration of overlap? | Reason for administration |
|---|---|---|---|---|
| 1 | Firocoxib 4.4 mg/kg/d PO, for 21 d | Meloxicam, unknown dose, SC, for 1 d | 1 d of overlap | Firocoxib use for rhinitis, meloxicam injection given by rDVM on day of ulcer diagnosis |
| 2 | Carprofen 4.8 mg/kg/d PO as needed, for 47 d | Meloxicam, 0.1 mg/kg, SC, for 6 d | No washout, but no overlap | Chronic use of carprofen for OA, meloxicam given in hospital during period of anorexia for treatment of limb swelling following mast cell tumor removal |
| 3 | Meloxicam unknown dose PO, for 5 d | Owner had given ibuprofen (65 mg/kg/d, PO, for 5 d, ending 2 wk before meloxicam prescription) | 2-wk washout | Both given for back pain |
| 4 | Deracoxib 2 mg/kg/d PO, for 7 d | Carprofen, 4.1 mg/kg, SC, during deracoxib course | 1 d of overlap | Carprofen injection given following laceration repair, deracoxib sent home |
| 5 | Carprofen 4.1 mg/kg/d PO, for 2 d | Deracoxib, 1.8 mg/kg/d, for 2 mo | 3-d washout with misoprostol administration during washout | Chronic deracoxib use for osteosarcoma, carprofen prescribed following amputation surgery |
| 6 | Prednisone 1 mg/kg/d PO, for 7 d | Carprofen, unknown dose, chronically (years) | 3 d of overlap | Chronic carprofen administration for OA, prednisone prescribed for unknown reasons |
rDVM = Referring veterinarian.
Ulcer diagnosis
Definitive diagnosis of a GI ulcer was attained with surgery in 21 dogs, necropsy in 8 dogs, and endoscopy in 4 dogs. A single ulcer was described in 19 dogs, and multiple ulcers were described in 14 dogs; only dogs that underwent necropsy or endoscopy were diagnosed with multiple ulcers. Overall, ulcer locations included the duodenum (n = 18), pylorus (11), greater curvature of the stomach (7), lesser curvature of the stomach (3), and distal esophagus (1). At least 1 ulcer was full thickness in 22 dogs based on surgical findings (n = 18) or necropsy (4). In the 22 dogs with full-thickness ulceration and septic peritonitis, ulcer location was the proximal duodenum (n = 13), pylorus (7), duodenal flexure (1), and greater curvature of the stomach (1).
Outcome following surgical diagnosis
In the 21 dogs that underwent surgery, procedures to address the ulcer included debridement with primary closure (n = 9) and partial gastrectomy (5), and 1 dog each received a duodenal resection and anastomosis, pyloroplasty with serosal patch placement, pylorectomy (gastroduodenal anastomosis), and ulcer debridement and closure with serosal patch placement. Closed suction drains were placed in the abdomen of 11 dogs. Two dogs were reported to have hemorrhage from the ulceration site, necessitating an intraoperative blood transfusion in 1 dog. Three dogs were euthanized on the table; 2 dogs with pyloric perforations were euthanized because of the ulcer, and 1 dog was euthanized because of pancreatic neoplasia identified during exploration. One dog arrested at the time of extubation.
In the perioperative period, packed RBC transfusions were administered to 5 dogs, fresh frozen plasma was administered to 5 dogs, fresh whole blood was administered to 3 dogs, and canine albumin was administered to 3 dogs. Postoperative medical management for the ulceration consisted of maropitant (n = 17), pantoprazole (9), sucralfate (8), omeprazole (4), misoprostol (3), ondansetron (3), famotidine (1), and ranitidine (1). Fourteen dogs received postoperative antibiotics, including ampicillin/sulbactam (n = 13), enrofloxacin (10), metronidazole (3), and clindamycin (1). Intraoperative postlavage culture results were available in 13 dogs, yielding no growth in 5 dogs, multiple species of bacteria in 3 dogs, Escherichia coli in 2 dogs, and Weisella, Clostridium, and Staphylococcus spp in 1 dog each (Table 4). Histopathology of the ulcerated region was performed in 7 dogs, with no neoplastic cells identified in any dog. The most common description included fibrinosuppurative enteritis and neutrophilic vasculitis. The dog euthanized due to a pancreatic mass had histopathology of the pancreas performed, which revealed a pancreatic endocrine carcinoma with < 1 mitotic figure/10 hpf. The ulcerated duodenal lesion was separate from the pancreatic mass in this dog.
The intraoperative culture of 13 dogs with surgically addressed gastrointestinal ulceration from NSAID or nonphysiologic doses of steroids.
| Dog | Bacterial isolate |
|---|---|
| 1 | Escherichia coli—hemolytic |
| 2 | Weissella sp |
| 4 | No growth |
| 5 | E coli, Lactobacillus sp, Staphylococcus epidermidis, Bacteroides thetaiotaomicron |
| 13 | No growth |
| 14 | Bacillus sp, E coli |
| 15 | Clostridium bifermentans, Lysinibacillus fusiformis, E coli |
| 17 | No growth |
| 20 | Clostridium perfringens |
| 22 | E coli—hemolytic |
| 23 | No growth |
| 24 | No growth |
| 26 | Nonhemolytic Staphylococcus sp |
The number for each dog corresponds to Table 1.
Median duration of hospitalization in the 17 dogs that survived the immediate postoperative period was 5.5 days (range, 1 to 29 days). Postoperative complications encountered included GI ileus (n = 5), hypokalemia (3), hypoalbuminemia (3), aspiration pneumonia (2), acute kidney injury (2), worsening anemia (2), pancreatitis (2), cardiac arrest (2), acute respiratory distress syndrome (1), hypercoagulability (1), hepatopathy (1), diarrhea (1), incisional site infection (1), disseminated intravascular coagulation (1), and jejunal intussusception (1), with some dogs having more than 1 complication. Thirteen out of the 21 dogs (61.9%) that underwent surgery survived until discharge. In addition to the 4 dogs that were euthanized or died in the immediate operative period, 3 dogs were euthanized due to poor prognosis a median of 3 days (range, 1 to 4.5 days) postoperatively, and 1 dog died via cardiac arrest < 24 hours following surgery.
Follow-up was available in 10 of the 13 dogs that survived to hospital discharge, with a median of 63 days (range, 21 to 492 days). Clinical signs were documented as improved or improving in 5 dogs on recheck examination. Status of clinical signs was unknown in the remaining dogs at the time of follow-up. One dog was euthanized 48 days after hospital discharge for intractable pain related to a lytic bone lesion. No dog was reported to return to the hospital for treatment of additional GI ulceration or recurrence of septic peritonitis.
Outcome following endoscopic diagnosis
Four dogs were diagnosed with GI ulceration using endoscopy. Two of these dogs had major concurrent diseases being treated at the time of ulcer diagnosis, including immune-mediated hemolytic anemia and active diabetic ketoacidosis. All dogs were treated medically with no surgical treatment pursued. Medical therapies for the ulcer included maropitant (n = 4), pantoprazole (3), omeprazole (2), ondansetron (1), and sucralfate (1). No blood products were administered. Three dogs were hospitalized for treatment for a median of 2 days (range, 1 to 5 days). One dog was treated on an outpatient basis only.
Histopathology was performed in all dogs that underwent diagnosis via endoscopy. Diagnoses were chronic gastritis and ulceration, chronic ulceration with neutrophilic inflammation, moderate to marked lymphoplasmacytic enteritis, and multifocal gastric erosions with superficial rod-shaped bacteria and fibrosis.
All dogs survived to hospital discharge; however, 1 dog was euthanized 2 days later due to lack of rapid improvement in clinical signs coupled with a several-month history of inappetence and lethargy. In the remaining 3 dogs, a median of 14 days (range, 7 to 28 days) of follow-up was available. Clinical signs related to the ulcer were reported to have improved in 1 dog, resolved in 1 dog, and were unknown in 1 dog.
Dogs with necropsy diagnosis
Ulceration was diagnosed on necropsy in 8 dogs. Medical management was initiated in 5 dogs for a median of 1 day (range, 1 to 2 days) prior to death or euthanasia, while 3 dogs were euthanized before intervention. Medical management included maropitant (n = 4), pantoprazole (4), sucralfate (1), ondansetron (1), ranitidine (1), and misoprostol (1). One dog received a fresh frozen plasma transfusion and packed RBC transfusion.
Seven dogs were euthanized due to poor prognosis, and 1 dog died via cardiac arrest secondary to hemorrhagic shock. All necropsies confirmed GI ulceration. Full-thickness ulceration with septic peritonitis was identified in 4 dogs. Other diagnoses found during necropsy included adrenocortical carcinoma and septic pleuropneumonia, renal papillary necrosis and intervertebral disk disease, widespread petechiae and ecchymosis due to presumed coagulopathy, disseminated intravascular coagulation, subcutaneous masses of the left foreleg and shoulder (presumed mast cell tumors), and end-stage liver cirrhosis with multiple acquired portosystemic shunts.
Discussion
In this study population, the most common NSAID to cause GI ulceration was carprofen, with 7 of the 13 dogs receiving the recommended dose with no concurrent NSAIDs or steroids. While subclinical ulceration has previously been described at therapeutic carprofen doses, this is the first report describing clinical ulceration at therapeutic carprofen doses, including 5 perforated ulcers.4 Meloxicam and deracoxib were the next most common NSAIDs used in the dogs of this study, which have both previously been associated with clinical GI ulceration.5 Full-thickness ulceration with secondary septic peritonitis was common in this population of dogs, and abdominal ultrasound was uncommonly successful at identifying the location of ulcerative lesions.
Previous reports suggest that appropriate NSAID usage typically does not result in ulceration; however, exceeding the recommended dose, combining NSAIDS, or combining an NSAID with a steroid can lead to ulceration.16,17 In our study, 7 dogs receiving carprofen alone at the recommended dose had ulcerations. Therefore, while it may be true that ulceration is more likely to occur when an NSAID is not used as recommended, it can also be observed when given appropriately. In several studies,11,18 carprofen resulted in subclinical GI ulceration when given short term and long term. Carprofen was the most common NSAID associated with ulceration in our study; however, this incidence may be increased since it is likely prescribed at a higher rate than other NSAIDs. This likely is not reflective of a higher risk profile, as carprofen has previously been shown to cause less severe subclinical GI lesions compared to other NSAIDs.11,18 It also has not been evaluated whether NSAIDs should be dosed based on lean body weight. The body condition score of the dogs in this study is unknown, and dogs receiving the recommended dose of NSAIDs that resulted in ulceration may not have been dosed based on their lean body weight. It has been found that there is a varied incidence of obesity in our canine populations, but dogs with certain diseases, such as orthopedic disease, are at higher risk for obesity, which is the most common reason for short- and long-term administration of NSAIDs.19 Therefore, further evaluation of NSAID dosing and body condition score and their effects on GI ulceration is needed.
Five dogs were receiving 2 NSAIDs concurrently, and 1 dog was receiving an NSAID and a steroid concurrently. All dogs on 2 NSAIDs concurrently experienced full-thickness perforation, whereas the dog on an NSAID and a steroid had partial-thickness ulceration. Concurrent administration of 2 NSAIDs may pose a greater risk of severe GI ulceration and septic peritonitis. Among the dogs with concurrent diseases that could potentially cause GI ulceration, one had a protein-losing enteropathy, and another had mast cell tumors of the elbow and scapula. The dog with protein-losing enteropathy was on prednisone at an anti-inflammatory dose. While high doses of corticosteroids can cause gastritis, they are unlikely to cause ulceration without an underlying condition, which may be the case here. The dog with mast cell tumors was receiving both meloxicam and carprofen concurrently, with carprofen administered above the recommended dosage. Therefore, it is difficult to determine whether the mast cell tumors contributed to development of ulceration. Similarly, several dogs were receiving additional medications; however, none of these medications have been previously associated with GI ulceration. Many dogs in this study also had comorbidities that have not been previously associated with GI ulceration; however, dogs in chronic pain or under distress have been found to have a higher incidence of GI ulceration.1 These comorbidities may have impacted the outcomes of the dogs in this study.
Abdominal ultrasound was performed in 24 of the 33 dogs (72.7%), with peritoneal effusion being the most common finding (n = 18 [75%]), followed by thickening of a portion of the GI tract (13 [54%]). These findings were consistent with those of previous studies,5,14 where peritoneal effusion and mural thickening are often associated with GI ulceration. However, these findings can occur with or without ulcer perforation and are relatively nonspecific. Peritoneal effusion is more commonly seen with GI perforation than without.14 In our study, the site of ulceration was identified in only 5 out of the 24 dogs (20.8%) with abdominal ultrasound, which aligns with previous reports indicating difficulty identifying the site of ulceration with abdominal ultrasound. In 1 study,14 CT was able to identify the site of ulceration in 79% of cases with perforated GI ulcers, compared to 14% with ultrasound. Considering these findings, CT could be considered in dogs suspected of having a perforated GI ulcer, as identifying the ulcer location can influence surgical planning and prognosis.
Of the 22 dogs diagnosed with septic peritonitis, 13 dogs had abdominocentesis performed prior to surgery or euthanasia. Intracellular bacteria were observed in only 4 out of the 13 samples (30%), with no bacteria seen in 7 samples and 2 samples deemed suspicious for bacteria. This is lower than the rates reported previously, where intracellular bacteria were observed in 56% to 72% of cases with septic peritonitis.5,14 The low number of intracellular bacteria observed in our study could be attributed to the low bacterial counts associated with proximal perforations or potential reduction of bacteria due to therapeutic interventions such as antibiotic therapy. Regardless, this finding highlighted the need to avoid relying on intracellular bacteria to confirm septic peritonitis in cases where perforation secondary to an upper GI ulcer is suspected. Patient history, other laboratory parameters, and imaging should be assessed for signs of sepsis or perforation to make a timely diagnosis.14,20,21
The most common site of ulceration in these dogs was the duodenum, observed in 18 dogs (54.5%), followed by the pylorus in 11 dogs (33.3%). Traditionally, it was believed that NSAID administration primarily leads to ulceration in the pylorus, followed by the proximal duodenum.3,11,13,14,17 However, in our study, there was a slightly higher rate of ulcerations occurring in the duodenum. These areas were also the most frequent locations for full-thickness ulceration, with 13 duodenal perforations (59.1%) and 7 pyloric perforations (31.8%). These findings were consistent with those of other studies.14 Localization of ulceration to the stomach and/or proximal duodenum is likely more prevalent due to increased gastric acid exposure to these portions of the GI tract, in combination with predisposing comorbidities or medications.
The main limitations of this study were its relatively small sample size and retrospective nature. The small sample size in this study likely reflected the inclusion criteria, which only allowed for inclusion of dogs with documented ulceration and likely excluded dogs with partial-thickness or subclinical ulceration for which definitive diagnosis was not pursued. Further prospective studies are needed to identify dogs with mild or subclinical ulceration and determine what causes dogs to progress to full-thickness ulceration. This also led to a low number of dogs receiving only a single steroid, affecting conclusion for steroid-administered dogs. Further research is needed to identify risk factors associated with steroid-related GI ulceration due to conflicting literature on its role in ulceration. Like many multi-institutional retrospective studies, there was no standardization of medical records, leading to missing data points and diagnostics. Additionally, variations in NSAID preferences among institutions and prescribing doctors may have influenced our results. We also did not evaluate whether dogs were dosed based on lean body weight. Thus, dogs with ulceration at appropriate doses may have been receiving a higher dose than if they had been dosed based on their lean body weight. Dogs with concurrent diseases contributing to GI ulceration were included, and these cases were documented.
In conclusion, our findings suggested that dogs are at risk for severe GI ulceration even when using COX-2–sparing NSAIDs at recommended doses. Concurrent use of another NSAID or steroid may increase the odds and severity of ulceration. Therefore, careful use and monitoring of dogs receiving NSAIDs, whether short or long term, should be considered. These medications should also be cautiously used in dogs with concurrent diseases known to cause GI ulceration.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org.
Acknowledgments
None reported.
Disclosures
The authors have nothing to disclose. No AI-assisted technologies were used in the generation of this manuscript.
Funding
The authors have nothing to disclose.
ORCID
V. Dickerson https://orcid.org/0009-0008-0096-9942
References
- 1.↑
Tolbert K, Gould E. Gastritis and gastric ulceration in dogs and cats. In: Bruyette DS, Bexfield N, Chretin JD, et al., eds. Clinical Small Animal Internal Medicine. Wiley; 2020:547-555. doi:10.1002/9781119501237.ch51
- 2.
Parrah J, Moulvi B, Gazi M, et al. Gastric ulceration in dog: a review. Vet World. 2013;6(7):449-454. doi:10.5455/vetworld.2013.449-454
- 3.↑
Stanton ME, Bright RM. Gastroduodenal ulceration in dogs: retrospective study of 43 cases and literature review. J Vet Intern Med. 1989;3(4):238-244. doi:10.1111/j.1939-1676.1989.tb00863.x
- 4.↑
Mabry K, Hill T, Tolbert MK. Prevalence of gastrointestinal lesions in dogs chronically treated with nonsteroidal anti-inflammatory drugs. J Vet Intern Med. 2021;35(2):853-859. doi:10.1111/jvim.16057
- 5.↑
Dobberstein REA, Oblak ML, Brisson BA, Singh A, Moens NMM, Ayoub S. Primary repair of nonsteroidal anti-inflammatory drug-associated full thickness gastrointestinal ulcers in 11 dogs. Vet Surg. 2022;51(7):1096-1105. doi:10.1111/vsu.13853
- 6.↑
KuKanich B, Bidgood T, Knesl O. Clinical pharmacology of nonsteroidal anti-inflammatory drugs in dogs. Vet Anaesth Analg. 2012;39(1):69-90. doi:10.1111/j.1467-2995.2011.00675.x
- 7.↑
Enberg TB, Braun LD, Kuzma AB. Gastrointestinal perforation in five dogs associated with the administration of meloxicam. J Vet Emerg Crit Care (San Antonio). 2006;16(1):34-43. doi:10.1111/j.1476-4431.2005.00157.x
- 8.↑
Case JB, Fick JL, Rooney MB. Proximal duodenal perforation in three dogs following deracoxib administration. J Am Anim Hosp Assoc. 2010;46(4):255-258. doi:10.5326/0460255
- 9.↑
Rohrer CR, Hill RC, Fischer A, et al. Gastric hemorrhage in dogs given high doses of methylprednisolone sodium succinate. Am J Vet Res. 1999;60(8):977-981. doi:10.2460/ajvr.1999.60.08.977
- 10.↑
Whittemore JC, Mooney AP, Price JM, Thomason J. Clinical, clinicopathologic, and gastrointestinal changes from aspirin, prednisone, or combination treatment in healthy research dogs: A double-blind randomized trial. J Vet Intern Med. 2019;33(5):1977-1987. doi:10.1111/jvim.15577
- 11.↑
Forsyth SF, Guilford WG, Haslett SJ, Godfrey J. Endoscopy of the gastroduodenal mucosa after carprofen, meloxicam and ketoprofen administration in dogs. J Small Anim Pract. 1998;39(9):421-424. doi:10.1111/j.1748-5827.1998.tb03748.x
- 12.↑
Whittemore JC, Mooney AP, Price JM, Thomason J. Clinical, clinicopathologic, and gastrointestinal changes from administration of clopidogrel, prednisone, or combination in healthy dogs: a double-blind randomized trial. J Vet Intern Med. 2019;33(6):2618-2627. doi:10.1111/jvim.15630
- 13.↑
Pavlova E, Gold RM, Tolbert MK, Lidbury JA. Medical conditions associated with gastroduodenal ulceration or erosion in 168 dogs: 2008-2018. J Vet Intern Med. 2021;35(6):2697-2704. doi:10.1111/jvim.16275
- 14.↑
Fitzgerald E, Barfield D, Lee KCL, Lamb CR. Clinical findings and results of diagnostic imaging in 82 dogs with gastrointestinal ulceration. J Small Anim Pract. 2017;58(4):211-218. doi:10.1111/jsap.12631
- 15.↑
Vetter TR. Fundamentals of research data and variables: the devil is in the details. Anesth Analg. 2017;125(4):1375-1380. doi:10.1213/ANE.0000000000002370
- 16.↑
Monteiro-Steagall BP, Steagall PVM, Lascelles BDX. Systematic review of nonsteroidal anti-inflammatory drug-induced adverse effects in dogs. J Vet Intern Med. 2013;27(5):1011-1019. doi:10.1111/jvim.12127
- 17.↑
Boston SE, Moens NMM, Kruth SA, Southorn EP. Endoscopic evaluation of the gastroduodenal mucosa to determine the safety of short-term concurrent administration of meloxicam and dexamethasone in healthy dogs. Am J Vet Res. 2003;64(11):1369-1375. doi:10.2460/ajvr.2003.64.1369
- 18.↑
Luna SPL, Basílio AC, Steagall PVM, et al. Evaluation of adverse effects of long-term oral administration of carprofen, etodolac, flunixin meglumine, ketoprofen, and meloxicam in dogs. Am J Vet Res. 2007;68(3):258-264. doi:10.2460/ajvr.68.3.258
- 19.↑
Chiang CF, Villaverde C, Chang WC, Fascetti AJ, Larsen JA. Prevalence, risk factors, and disease associations of overweight and obesity in dogs that visited the veterinary medical teaching hospital at the University of California, Davis from January 2006 to December 2015. Top Companion Anim Med. 2022;48:100640. doi:10.1016/j.tcam.2022.100640
- 20.↑
Ragetly GR, Bennett RA, Ragetly CA. Septic peritonitis: etiology, pathophysiology, and diagnosis. Compend Contin Educ Vet. 2011;33(10):E1-E6.
- 21.↑
Ben Oz J, Aroch I, Segev G. Increased ratio of peritoneal effusion-to-serum potassium concentration in a dog with gastric perforation. J Vet Emerg Crit Care (San Antonio). 2016;26(6):793-797. doi:10.1111/vec.12453
