Gastric perforation secondary to incarcerated para-esophageal hiatal hernia in a French bulldog

Abstract

A 6-month-old, intact female, French bulldog was presented to the Emergency Department for evaluation of vomiting and diarrhea over the preceding week which had not responded to supportive medical therapy. Imaging studies identified an incarcerated para-esophageal hernia with peritoneal effusion and gas consistent with gastrointestinal perforation. Following stabilization, the dog underwent an exploratory laparotomy which confirmed an incarcerated hiatal hernia and gastric perforation. A gastrectomy was performed to repair the defect, and to prevent recurrence both a herniorrhaphy and esophagopexy were performed. Post-operative care required treatment for septic shock including vasopressor and hydrocortisone infusions and plasma transfusions for colloidal support. The patient was successfully discharged 4 days after surgery. The histopathology results identified spiral bacteria consistent with Helicobacter spp. which was subsequently treated with oral antibiotics and a proton pump inhibitor. The dog has had no further gastrointestinal signs in the 90 days since surgery.

Gastric perforation and peritonitis can occur secondary to an incarcerated esophageal hiatal hernia, and if treated promptly can result in a successful outcome. This case demonstrates a novel etiology of gastric perforation which may be associated with brachycephalic breeds.

Hiatal hernias are characterized by displacement or projection of abdominal organs or structures into the mediastinum through the hiatus of the diaphragm where the esophagus joins the stomach (1–3). Hiatal hernias are classified as:

• Type I (sliding hiatal hernia) — cranial displacement of the abdominal esophagus, esophagogastric junction with or without the stomach;

• Type II (paraesophageal hernia) — involves the stomach herniating through the esophageal hiatus while the gastroesophageal junction remains correctly positioned below the diaphragm;

• Type III — a mixed hernia with a sliding and paraesophageal component; and

• Type IV — herniation of abdominal organs into the thoracic cavity (2,4).

Although uncommon in the overall canine population, hiatal hernias are becoming increasingly more recognized in brachycephalic breeds (3).

Type I hernias can be medically or surgically managed depending on severity of associated clinical signs, whereas surgical intervention is recommended for Type II hiatal hernias due to the complications associated with an abnormally placed intrathoracic stomach (2). In humans, the reported complications include volvulus, obstruction, incarceration, and perforation, all of which are considered rare presentations for acute abdomen (5). Medical and surgical management of Types I and II hiatal hernias to prevent associated clinical signs are well-described in brachycephalic dogs; however, gastric perforation has not previously been reported as a complication of hiatal hernias in dogs. The purpose of this report is to describe the successful management of gastric perforation and peritonitis secondary to an incarcerated paraesophageal hernia in a brachycephalic breed.

Case description

A 6-month-old, 10 kg, intact female, French bulldog was presented to a private referral hospital emergency department for evaluation of vomiting and diarrhea lasting 7 d. The day prior the dog was evaluated by a primary care veterinarian for vomiting and anorexia and was treated as an outpatient with maropitant [Zoetis, Rhodes, New South Wales (NSW), Australia]. The owners had not reported any clinical signs associated with gastrointestinal disease prior.

On presentation the dog was quiet, but alert. The heart rate was 160 beats/min, with normal pulse quality, mean arterial pressure of 85 mmHg, pale pink mucous membranes with a capillary refill time of 2 s. There was cranial abdominal pain and melena on rectal examination; core temperature was 38.1°C; there were no other significant findings on examination. The dog was given methadone (Ilium, Sydney, NSW, Australia), 0.2 mg/kg, IV and a 10 mL/kg bolus of a balanced crystalloid solution (Hartmanns; Baxter, Old Toongabbie, Sydney, NSW, Australia) which was then continued at 4 mL/kg per hour.

A complete blood (cell) count showed a marked reticulocytosis [416 800 cells/μL; reference range (RR): 10 000 to 110 000 cells/μL] with a hemotocrit of 0.54%. A differential white blood count showed neutropenia (2320 cells/μL; RR: 2950 to 11 640 cells/μL), lymphocytosis (5120 cells/μL; RR: 1050 to 5100 cells/μL), monocytosis (1700 cells/μL; RR: 160 to 1120 cells/μL), eosinopenia (20 cells/μL; RR: 600 to 1230 cells/μL) with a normal platelet count (481 000/μL; RR: 148 000 to 484 000/μL); all other parameters were within reference ranges. Serum biochemistry showed a mild elevation in urea (9.7 mmol/L; RR: 2.5 to 9.6 mmol/L) and hyperphosphatemia (2.61 mmol/L; RR: 0.81 to 2.20 mmol/L), total protein (56 g/L; RR: 55 to 75 g/L); all other parameters were within reference ranges. Venous blood gas showed a mixed respiratory and metabolic acidosis (pH 7.283; RR: 7.35 to 7.46), base excess (−4.0 mmol/L; RR: −2 to +2 mmol/L), PvCO₂ (47.9 mmHg; RR 32 to 43 mmHg), hypokalemia (3.5 mmol/L; RR: 3.9 to 4.5 mmol/L), hyponatremia (138 mmol/L; RR: 140 to 150 mmol/L), hypochloridemia (101 mmol/L; RR: 109 to 120), ionized hypocalcemia (1.22 mmol/L; RR: 1.29 to 1.5 mmol/L), and hyperlactatemia (6.1 mmol/L; RR: 0 to 2.5 mmol/L). Both 3-view thoracic and abdominal radiographs as well as an abdominal ultrasound were performed and reviewed by a Board-certified radiologist. Thoracic and abdominal radiographs showed a well-demarcated gas-filled soft tissue opacity within the caudal esophagus, loss of normal gastric anatomy within the cranial abdomen and a pneumoperitoneum (Figure 1). Abdominal ultrasound identified a dilated fluid-filled stomach, a very wide esophageal hiatus allowing peritoneal fat into the caudal mediastinum with part of the fundus, with a hyperechoic peritoneum and a marked volume of hypoechoic peritoneal fluid (Figure 2). Ultrasound-guided abdominocentesis was performed and cytology showed large numbers of neutrophils but no obvious microorganisms. Based on this, a diagnosis of peritonitis was made secondary to a para-esophageal herniation and perforation of the gastric fundus.

Figure 1.

Left recumbent lateral radiograph of the thorax and cranial abdomen. Moderate gas dilatation of the cranial and mid-thoracic esophagus. Smoothly rounded soft tissue opacity summating with the caudal thoracic esophagus. The cranial margin is well-defined, with gas/soft tissue interface (arrow). The pylorus (P), gastric body (B) are indicated. The gastric fundus (F) is cranially displaced within the caudodorsal thorax. Numerous foci of gas are identified in the peritoneal space (arrowheads). 203 × 148 mm (300 × 300 DPI).

Figure 2.

Transverse ultrasound image of the cranial abdomen, with patient’s right to the left of image, and dorsal aspect of the patient to the bottom of the image. The liver (Li) is to the patient’s right. The gastric pylorus (P), body (B) and fundus (F) are indicated. The diaphragmatic hiatus is identified by the black dotted line and is extremely widened. There is focal extension of peritoneal fat into the caudal mediastinum. On this current image the fundus is located ventral/caudal to the hiatus, but on dynamic images, variable extension into the caudal mediastinum was observed. The peritoneal fat is hyperechoic. 169 × 104 mm (300 × 300 DPI).

Repeat assessment of the dog identified worsening hemodynamic and perfusion parameters with prolonged capillary refill time, absent peripheral pulses, and no measurable blood pressure. Fluid resuscitation using Hartmann’s at a total dose of 30 mL/kg improved pulse quality and attained a measurable systolic blood pressure of 90 mmHg. The dog became refractory to further fluid boluses and was started on a norepinephrine (Pfizer, Sydney, NSW, Australia) infusion at 0.3 μg/kg per minute, IV. Due to concerns for gastric perforation and leakage of gastric contents, empirical antibiotics of amoxicillin-clavulanate (Juno, Cremorne, NSW, Australia) 20 mg/kg, IV, q8h and enrofloxacin (Ilium) 10 mg/kg IV, q24h were commenced. The dog’s American Society of Anesthesiologists (ASA) score at induction was 4E; defined as a patient with severe systemic disease that is a constant threat to life, where delay in treatment would lead to significant increase in the threat to life (6).

The dog was premedicated with methadone 0.1 mg/kg, IV and induced with fentanyl (AstraZeneca, North Ryde, NSW, Australia), 10 μg/kg, IV and midazolam (Pfizer), 0.3 mg/kg, IV, intubated and maintained on total intravenous anaesthesia with propofol (Sandoz, Macquarie Park, NSW, Australia), 0.2 μg/kg per minute, IV and fentanyl, 5 μg/kg per minute, IV. The dog was persistently hypotensive after induction of anesthesia resulting in norepinephrine being increased to 1.5 μg/kg per minute, IV and the addition of a dobutamine (Hospira, Melbourne, Victoria, Australia) infusion at 10 μg/kg per minute, IV to improve cardiac inotropy. A midline laparotomy and exploration of the abdomen was performed. Upon entering the abdominal cavity approximately 500 mL of free abdominal fluid was encountered, as well as a large full thickness defect at the gastric fundus near the esophagus. This was associated with a significantly lax esophageal hiatus allowing the passage of 3 fingers into the thoracic cavity. The malpositioned stomach was replaced into the abdominal cavity and a partial gastrectomy was performed to remove abnormal tissue secondary to the incarcerated paraesophageal hernia. No other abnormalities were detected in the stomach, the remainder of the gastrointestinal tract, or other abdominal organs. The stomach was routinely closed in 2 layers and a herniorrhaphy and esophagopexy were performed around the hiatus using 3-0 polydioxanone suture material (PDS*II, Ethicon; J&J Medical, North Ryde, NSW, Australia) in an interrupted pattern. After copious peritoneal lavage, Jackson Pratt drains were placed prior to routine 3-layer abdominal closure. The peritoneal fluid was submitted for culture and susceptibility, and the resected portion of stomach was submitted for histopathology.

Post-operatively the dog was hypothermic, hypotensive, and stuporous despite further crystalloid boluses and high infusion rates of norepinephrine at 1.5 μg/kg per minute and dobutamine at 10 μg/kg per minute. Additional medications commenced immediately after surgery included hydrocortisone (Pfizer) at 0.2 mg/kg per hour, IV, dalteparin (Pfizer), 150 IU/kg, SC, q8h, esomeprazole (AstraZeneca),1 mg/kg, IV, q12h, and paracetamol (Pfizer), 10 mg/kg, IV, q12h. During the first 6 h after surgery, both norepinephrine and dobutamine were weaned and discontinued as the dog’s blood pressure and perfusion status improved. Day 1 after surgery, the dog was again becoming hemodynamically unstable defined by progressive hypotension and tachycardia. At this stage the dog was also profoundly hypoproteinaemic with a total protein (28 g/L; RR: 55 to 75 g/L) and showed signs of volume overload with peripheral edema, most notable in the conjunctiva and a weight gain of 10% body weight. These findings raised concerns for increased vascular permeability from endothelial glycocalyx damage with resultant large volumes of crystalloid therapy accumulating in the interstitial space. A fresh frozen plasma (PLASVACC, Kabar, Queensland, Australia) transfusion 10 mL/kg, IV over 2 h was administered to provide colloidal and endothelial glycocalyx support. By completion of the transfusion the dog’s perfusion parameters improved and lactate normalized, and over the following 24 h the peripheral edema reduced, evident by reduction in weight and reduced conjunctival swelling.

On Day 2 after surgery, the dog became anemic with a PCV of 24%, a decline from 34% the day prior with no change in hemodynamic stability; therefore, no blood products were administered. On Day 3 after surgery, the dog’s PCV had improved to 30% suggestive of no further hemorrhage and active regeneration. The dog continued to improve clinically with no evidence of surgical dehiscence based on daily drain fluid analysis. Prior to discharge interim culture results showed no anerobic or aerobic growth; therefore, antimicrobials were de-escalated. The patient was discharged 4 d after surgery with the following medications: enrofloxacin 10 mg/kg, PO, q24h for 7 d, omeprazole 1 mg/kg, PO, q12h for 10 d, maropitant 2 mg/kg, PO, q24h for 4 d, and paracetamol 10 mg/kg, PO, q12h for 5 d. Ten days after discharge the dog was clinically well; however, there was a focal area of dehiscence of the surgical skin incision and associated purulent discharge that was treated with amoxicillin-clavulanate 20 mg/kg, PO, q12h for a further 7 d.

The final culture results from the peritoneal fluid had no microbial growth. The histopathology results identified a tract lined by a marked exudate of hemorrhage, fibrin, and neutrophils extending through gastric mucosa, submucosa, and muscularis, bordered by markedly edematous, proliferating fibrovascular tissue with subacute focal perforated gastric ulcer, peritonitis, and moderate numbers of large spiral bacteria in the surface mucus and superficial glands consistent with Helicobacter spp. Treatment for pathologic Helicobacter spp. infection with metronidazole (Sanofi-Aventis, Macquarie Park, NSW, Australia), 10 mg/kg, PO, q12h and amoxicillin-clavulanate 20 mg/kg, PO, q12h and omeprazole 1 mg/kg, PO, q12h were continued for a total of 2 wk. Three months after surgery the dog has made a full recovery with no further gastrointestinal signs.

Discussion

To the authors’ knowledge this is the first report of gastric necrosis and rupture associated with an incarcerated hiatal hernia. Given the breed and surgical findings, it is likely that this dog had a pre-existing Type II hiatal hernia that became incarcerated due to the malpositioned intrathoracic stomach and developed ischemic necrosis due to focal impedance of gastric perfusion (5,7). The dog had acute gastrointestinal signs lasting 7 d with progression to hematemesis suggesting gastric ulceration development during this timeframe and ensuing perforation due to secondary ischemic injury and resultant peritonitis. Strangulation of a hiatal hernia and ischemic injury leading to perforation is a reported cause of gastric perforation in humans (7). As the stomach is fixed at the gastroesophageal junction, migration of the stomach into the thorax can result in rotation around its longitudinal axis resulting in volvulus, obstruction, incarceration, or gastric ulceration and bleeding (5). In humans, perforation secondary to gastric ischemia is the main cause of mortality, whereas the incidence of acute complications due to volvulus or incarceration are rare, with an incidence of only 1.2% per patient per year (8). Unlike Type I hiatal hernia, Type II hiatal hernia management is generally surgical, to either prevent catastrophic complications, or as in this case treat the acute abdomen (1,2). Type II hiatal hernias retain normal lower esophageal sphincter function with the signs related to the gastric herniation through the enlarged hiatus in the diaphragm. Surgery primarily aims at maintaining reduction and closure of the hiatal hernia (2,9).

With the rising popularity of brachycephalic breeds there is increasing literature regarding the high prevalence of hiatal hernias (3). In French bulldogs, up to 76% presenting for brachycephalic airway surgery evaluation are diagnosed with a Type I hiatal hernia (3). An increase in inspiratory effort, and therefore increase in negative pressure within the upper airways, is believed to predispose brachycephalic breeds to both digestive tract and respiratory abnormalities (10,11). The consequences of increased negative pressure within the upper respiratory tract leads to increased intraesophageal and intrapleural pressure which can pull the stomach and esophagus into the thoracic cavity (2,11). Given the underlying pathophysiology for acquired hiatal hernias in French bulldogs, this dog’s surgical intervention included procedures to correct both Type I and II hiatal hernias; herniorrhaphy (closure of hiatal defect), and esophagopexy (to prevent future displacement). The partial gastrectomy was required to repair the gastric perforation; a known complication of Type II hernias in humans (8). A Type II hernia has also been documented in a single case report involving a Boston terrier which was also presented for acute abdominal signs and was diagnosed with gastric dilatation and volvulus (4). Like this dog, the presumption is that the underlying conformational disorder predisposed to a Type II hiatal hernia.

Prior to surgery, the dog was in a state of both hypovolemic and distributive shock with hypotension and persistent hyperlactatemia (5.7 mmol/L; RR: 0 to 2.5 mmol/L) which had become refractory to further fluid resuscitation, therefore requiring vasopressor therapy (12). Refractory shock secondary to a systemic inflammatory response syndrome is associated with greater risk of mortality due to profound circulatory, cellular, and metabolic abnormalities (13). Survival rates for peritonitis secondary to gastrointestinal perforation are reported from 27 to 85% with most approximating a 50 to 70% survival (14,15). Risk factors for mortality in dogs with preoperative peritonitis include pre- and post-operative hypoproteinemia, intraoperative hypotension, lower preoperative red blood cell count, and higher white blood cell count, all of which are suggestive of severe systemic disease (16). Other studies have also shown that those with an ASA score ≥ III are more likely to dehisce post-operatively, whereas this and higher pre-operative lactate are associated with increased mortality (17). This dog had multiple pre- and postoperative risk factors for mortality, including preoperative peritonitis and hyperlactatemia, an ASA grade of IVE and post-operative hypoproteinemia and hypotension. Within the first 12 h after surgery, the dog became profoundly hypoproteinemic (TP 28g/L; RR: 55 to 75 g/L) with peripheral edema affecting all limbs and the ocular conjunctiva. The source of edema and hypoproteinemia was thought to be due to a combination of increased vascular permeability, negative acute phase response to inflammation, dilution from crystalloid fluid therapy, and third space fluid losses into interstitium and peritoneal cavity. In previous veterinary and human studies hypoalbuminemia is a predictor of morbidity and mortality (15,18). This is likely due to the important role of albumin in maintaining oncotic pressure, mediating inflammation and coagulation, protection from oxidative damage, and providing hormonal and drug binding capacity (15,19).

The progressive anemia was likely secondary to gastrointestinal hemorrhage given the presence of preoperative hematemesis, melena, and reticulocytosis indicating regeneration. In humans Cameron lesions are a unique consequence of all hiatal hernias (20). However, there is an association with size, with a higher prevalence in those larger than 5 cm (21). These lesions are a result of trauma and ischemia caused by the hiatal hernia and can result in either chronic blood loss or acute bleeding, reported in up to 13.7% patients (20,21). They are typically identified via endoscopy as linear gastric erosions or ulcers on the mucosal folds on the lesser curvature of the stomach at the level of the diaphragmatic hiatus (20,22). Such lesions have not yet been reported in dogs. A Cameron lesion may have been contributing to the ulceration and hemorrhage and should be considered and investigated as a cause of acute or chronic hemorrhage in patients with a hiatal hernia. Other documented causes of gastrointestinal ulceration include anti-inflammatory medications, foreign object ingestion, neoplasia, strenuous exercise, inflammatory bowel disease, hepatic disease, or uremia (23). These other causes were excluded in this dog based on history, diagnostics, and surgical findings.

Histopathology revealed a subacute focal perforated gastric ulcer with no evidence of neoplastic cells; there were moderate numbers of spiral bacteria morphologically consistent with Helicobacter spp. in the superficial glands. In humans, Helicobacter pylori is a common and important cause of gastritis and gastric ulceration, affecting approximately 4.4 billion individuals worldwide (24). Given the association of H. pylori with significant gastric disease, treatment in humans is generally recommended; however, there are increasing levels of antimicrobial resistance worldwide (24,25). In dogs, Helicobacter spp. treatment remains controversial primarily due to high prevalence of 67 to 100% in healthy dogs and unknown pathogenicity of the numerous different species (26). Generally, gastric Helicobacter spp. in dogs are larger than H. pylori; however, these are indistinguishable by routine light microscopy and multiple species may be present (27). In this case, it could not be determined whether the Helicobacter spp. were truly clinically significant, since the species was not identified. However, given that Helicobacter spp. can cause gastric ulceration and that no other infectious organisms were identified the decision was made to treat empirically. There are various treatment protocols but the overall guidelines from human recommendations are administration of antimicrobials effective against Helicobacter spp. and an acid-secretory inhibitory drug for 2 wk (27). Following identification of Helicobacter spp., therapy was instituted for a total of 2 wk.

The dog in this case likely had a pre-existing paraesophageal hernia, an anatomical abnormality associated with brachycephalic breeds. The significance of such abnormalities can become life-threatening due to complications associated with incarceration of the malpositioned stomach leading to either gastric dilatation and volvulus, chronic gastric bleeding, or gastric perforation. This case documents the successful outcome of a novel source of septic peritonitis secondary to a Type II hiatal hernia. In dogs with hiatal hernia, gastric perforation should be considered when presenting with acute and severe gastrointestinal signs as this requires emergent surgical management.