Endoscopic Diagnosis of a Pylorogastric Intussusception with Spontaneous Resolution^S

Introduction

An intussusception results from the invagination of a portion of the intestine (intussusceptum) into the lumen of an adjacent intestinal segment (intussuscepiens).^1–3 By convention, the invaginating segment is named first, followed by the recipient segment when referring to the type of intussusception.^1,2

Various types of gastrointestinal intussusception have been described in dogs and cats, including gastroesophageal, gastrogastric, pylorogastric, duodenogastric, enteroenteric, enterocolic (ileocolic), ileocecal, cecocolic, colocolic, and colorectal.^2–11 Intussusceptions occur most commonly at the ileocolic junction in young animals. There have only been three other reported cases of pylorogastric intussusceptions (PGIs) in dogs.^4,8,9 Other gastric outflow intussusceptions have been described in dogs and include one case each of gastroduodenal, gastrogastric, and duodenogastric intussusception.^5,6,11 Only two cases of duodenogastric intussusception and five of gastrogastric intussusception have been reported in humans.^12–16

Case Report

A 7 mo old intact female golden retriever weighing 21 kg presented to our Veterinary Teaching Hospital for an acute onset of vomiting. In the 2 hr preceding presentation, the patient had vomited clear fluid and mucus 14 times. The owners reported normal appetite, activity level, and water intake before the onset of vomiting. There was no history of medical illness nor had estrus yet been observed. The owner was questioned thoroughly regarding the possibility of foreign material ingestion, recent changes in diet or dietary indiscretion, toxin exposure, or any other potential cause for the acute vomiting, but an inciting cause was not apparent. The dog was well supervised and confined indoors except for leash walks and occasional access to a fenced yard. Additionally, the dog was routinely crated overnight. The dog's diet consisted of a commercial large breed puppy formula^a and occasional dog treats. The dog had been vaccinated appropriately with a total of four doses of distemper-adenovirus-parainfluenza-parvovirus^b vaccine and a single dose of rabies vaccine. For routine parasite control, pyrantel pamoate^c had been given 6 mo previously and milbemycin oxime^d had been given 4 and 3 mo before presentation. The current owner had acquired the dog from a kennel in California 4 mo earlier. Two additional healthy dogs resided in the household, including one dog acquired 4 days previously.

On physical examination, the dog was bright, alert, and responsive. Her body condition score was 3/5. She was estimated to be 6–7% dehydrated based on decreased skin turgor, dry mucous membranes, slow capillary refill time (2 sec), and tachycardia (heart rate, 160 beats/min). Pain and nausea could also have been contributing factors for the tachycardia. Systolic blood pressure was 130 mm Hg. Pulmonary auscultation was within normal limits. Abdominal palpation revealed a 2 × 4 cm firm, nonpainful, slightly movable structure in the midabdominal region, most likely to be the spleen. The dog vomited clear mucoid fluid three times during the physical examination.

Due to the findings on abdominal palpation and the history of acute vomiting, three-view abdominal radiographs were obtained. A 6 cm diameter, distinctly marginated, rounded soft tissue opacity suspected to be a foreign body was present within the stomach, best seen in the left lateral and ventro-dorsal projections (Figure 1). The small intestines were fluid filled but otherwise unremarkable with no evidence of obstruction. A large amount of normal-appearing fecal material was present throughout the colon. Serum electrolyte panel, packed red cell volume, and total serum protein were within normal limits.

View Figure

• Download as Powerpoint
• Download Figure

In light of the suspected gastric foreign body and profuse nature of the vomiting, an immediate exploratory laparotomy or upper gastrointestinal endoscopy was recommended to the owner. The owner opted for endoscopy first, followed by surgery if indicated. A peripheral intravenous catheter was placed, and Lactated Ringers Solution^e was given IV at 100 mL/hr (approximately 5 mL/kg/hr) for initial correction of dehydration. Anesthesia was induced with propofol^f (5.7 mg/kg IV), an endotracheal tube was placed, and anesthesia was maintained with sevoflurane^g.

On endoscopy^h, the only lesion noted in the esophagus was a small semicircular mucosal erosion adjacent to the gastroesophageal sphincter and mild dilation of the sphincter. Upon entering the stomach, a large tubular mass was noted to fill the pyloric antrum. The surface of this tubular mass appeared to be lined with edematous, folded mucosa that was contiguous with the normal mucosa of the gastric antrum at its base (see Supplementary Video 1). The mucosa of the tubular mass had multifocal superficial erosions and was covered with dark brown digested blood. The tubular structure was observed to repeatedly invaginate retrograde from the lumen of the antrum to the body of the stomach and then move anterograde, becoming partially reduced, back into the lumen of the antrum (see Supplementary Video 2). On circumferential evaluation of the base of the tubular structure, the mucosa was noted to be contiguous with the wall of the antrum. The pylorus was not visualized in its normal anatomic location. After extensive examination, it was determined that the tubular structure was an intussusceptum and the pyloric opening was located centrally in the apex of the intussusceptum. The unaffected body and fundic regions of the stomach contained a small amount of clear fluid and had normal appearing gastric mucosa. The dog was moved from left to right lateral recumbency to evaluate the gastric outflow tract from a different endoscopic perspective. The endoscopic findings in right lateral recumbency were similar and also consistent with a PGI. Manual reduction of the intussusception was attempted by pressing the tip of the endoscope against the tip of the intussusceptum. When that failed, exploratory laparotomy was recommended to further characterize and reduce the intussusception.

Surgery was performed immediately after the endoscopy under the same anesthesia. Hydromorphoneⁱ (0.05 mg/kg IV) and fentanyl^j (2 μg/kg/hr, IV by constant rate infusion) were given before surgery for presurgical pain management. The fentanyl infusion was continued during surgery and postoperatively. An exploratory laparotomy was performed, and the stomach was isolated and inspected. The serosal surfaces appeared to be within normal limits, and the anatomic relationship of the stomach, pylorus, and duodenum were normal. The PGI observed at endoscopy was no longer present. Upon palpation of the stomach, the gastric wall was found to be circumferentially firm and diffusely thickened to approximately 1.5 cm up to a distance of 5 cm orad from the pylorus. A gastrotomy incision was made between the greater and lesser curvature adjacent to the thickened area. Additionally, on digital palpation evaluation of the gastric outflow tract, there was no evidence of an intraluminal obstruction or the presence of the invaginating tubular structure (intussusceptum).

No significant abnormalities were noted in the remaining abdominal organs. A full thickness biopsy was obtained from the margin of the gastrotomy site, and a mucosal biopsy was taken of the thickened gastric wall. Impression smears that were taken of the gastric mucosa were stained with Romanowski stain^k, which revealed many spiral bacteria scattered throughout the slide. After closing the gastrotomy site, a routine ovariohysterectomy was performed. Anesthesia recovery was uneventful.

The dog was kept in the intensive care unit for 45 hr postoperatively. During this time, IV fluids were continued and the dog received maropitant^l (1 mg/kg subcutaneously q 24 hr), famotidine^m (0.5 mg/kg IV q 12 hr), sucralfateⁿ (50 mg/kg per os [PO] q 8 hr), and fentanyl^j (2 μg/kg/hr, IV by continuous rate infusion). The dog received nothing PO for the first 12 hr, followed by feeding of small meatball-sized portions of Prescription Diet I/D^o or EN Formula^p q 4–6 hr. Vomiting did not occur postoperatively, but one episode of regurgitation was observed 12 hr after surgery. The dog was discharged from the hospital on sucralfate (50 mg/kg PO q 8 hr), tramadol^q (2.4 mg/kg PO q 12 hr), and famotidine (0.5 mg/kg PO q 12 hr).

The full-thickness gastric biopsy revealed moderate mucosal congestion, severe mural (primarily serosal) edema, and numerous large spiral bacteria in the mucosal glands. These bacteria were presumably Helicobacter sp., but speciation was not performed. Additional findings included transmural moderate diffuse eosinophilic and neutrophilic gastritis and lamina propria lymphoid follicle formation (lymphoid follicular hypertrophy). The separate gastric mucosal biopsy showed the presence of more severe edema and congestion.

At re-evaluation and skin staple removal 9 days after surgery, the owners reported that no further vomiting had occurred, and the dog was clinically normal. Although it was uncertain as to what role, if any, the Helicobacter-like organisms contributed to the acute vomiting or the occurrence of pylorogastric intussusception, empirical treatment was instituted with a combination of clarithromycin^r (12 mg/kg PO q 12 hr for 21 days), amoxicillin^s (19 mg/kg PO q 12 hr for 21 days), metronidazole^t (12 mg/kg PO q 12 hr for 21 days), and omeprazole^u (1 mg/kg PO q 12 hr for 21 days). The last follow up was performed 6 mo after surgery, and no further vomiting had occurred. Further follow up regarding the Helicobacter status of this dog after treatment was not performed.

Discussion

Review of the veterinary literature suggested that gastrointestinal intussusceptions might be more common in young dogs. A retrospective study of 40 dogs and 5 cats with various types of gastrointestinal intussusception found that 37 of 45 (82.2%) patients were <1 yr of age.³ Similarly, children also appear to have a higher incidence of intussusceptions compared with adults.¹⁵ Although the number of reported canine cases of gastric outflow intussusception is low, five of seven cases (including the case reported here) occurred in dogs <4 yr of age.^4–6,8,9,11 Gastric outflow intussusception might also occur more frequently in large breed dogs as five of the seven reported cases (including the dog of this report) occurred in large breeds.^4,6,9,11 Although the paucity of case reports of gastric outflow intussusception did not allow definitive conclusions to be made regarding age and breed predilections, the dog reported here seemed to fit the typical profile of young age and large breed.

In humans, gastric outflow intussusception occurs most frequently in children (90%), and a definitive cause is only found in approximately 5% of them, which may include a polyp, Meckel's diverticulum, or indwelling gastrostomy tube.^13,15 Adults that presented with these types of intussusceptions usually had predisposing gastric neoplasia or occasionally polyps.^12,15,16

Intestinal intussusceptions are thought to occur as a result of a lack of homogeneity of the bowel wall, which may be caused by any abnormality within the bowel wall that alters local intestinal motility or pliability.¹ Most canine intussusceptions are idiopathic. Less commonly, intussusception occurs secondary to identifiable conditions of the gastrointestinal tract, including intestinal parasitism, gastroenteritis, intoxication, ingestion of bones, linear intestinal foreign bodies, parvovirus, distemper virus, intestinal neoplasia, granulomas (e.g., histoplasmosis, toxoplasmosis), and prior abdominal surgery.^3,17,18

Many of these causes were ruled out in this case. There was no evidence to support intoxication, dietary factors, foreign body, infectious enteritis, neoplasia, bone ingestion, prior abdominal surgery, or granuloma. Dietary intolerance or food allergy was also impossible because the dog recovered from surgery and remained asymptomatic for at least 6 mo (last follow up). The remaining potential causes included idiopathic intussusception, gastritis, or subclinical parasitism.

Of the three previously described dogs with PGI, two had a history of dietary indiscretion and profuse vomiting, which could have been predisposing factors to PGI.^8,9 Additionally, the only dog with duodenogastric intussusception reported in the veterinary literature also had a history of dietary indiscretion before presentation.⁵ To the best of our knowledge, this dog's history did not support dietary indiscretion. Subclinical parasitism could not be excluded as a possibility, especially considering the eosinophil component of the inflammatory response found in the gastric biopsies. Although it did not exclude the possibility of a recently acquired parasite infection, this dog had received a dose of pyrantel pamoate 6 mo previously and doses of milbemycin oxime 3 and 4 mo previously. Milbemycin oxime has been considered an effective adulticide for most hookworms and whipworms in the dog at single dosages although monthly treatment is typically recommended.¹⁹ However, in retrospect, it might have also been justified to also empirically treat this dog for subclinical gastrointestinal parasite infection.

This dog most likely had an underlying gastritis in addition to the PGI, as an intussusception of <24 hr duration would not be expected to cause diffuse transmural eosinophilic inflammation. Some authors suggested that acute gastritis or gastroenteritis is the most common predisposing factor for intestinal intussusception in young dogs where viral gastroenteritis is common.²⁰ However, others suggested that the majority of intestinal intussusceptions in dogs and cats are idiopathic, as in the case of this vaccinated dog.³ This dog had histopathological changes consistent with gastritis, which could have altered the gastric motility and been a risk factor. The excess number of spiral bacteria seen in some of the areas sampled, neutrophilic gastritis, and lymphoid follicle formation suggested that this dog might have Helicobacter gastritis. However, Helicobacter-like organisms frequently colonize the stomach in dogs without causing clinical signs; thus, it was difficult to implicate the Helicobacter-like organisms as the predisposing factor for this dog's PGI.²¹ The absence of clinical signs preceding the acute vomiting episodes and lack of data supporting a definitive pathologic role of Helicobacter sp. could have argued against treating this dog for a Helicobacter infection. However, the large number of spiral bacteria in association with unexplained gastritis and the absence of identifiable causes for the PGI led to the decision to empirically treat this dog for Helicobacter infection.

Similar to previous reports of PGI in dogs, vomiting was the presenting complaint in this case.^4–6,8,9 Vomiting might have been the underlying cause of the intussusception. Most intussusceptions occur aborally in the direction of peristalsis; however, it was possible in this case that reverse peristalsis might have played a role in the development of this dog's retrograde intussusception. Reverse peristalsis with retrograde movement of duodenal contents into the stomach is physiologically part of the vomiting reflex; therefore, profuse vomiting might have contributed to the PGI.² Alternatively, the acute vomiting might have been the consequence of the intussusception. In most reported cases of gastric outflow intussusception in dogs, preceding causes of vomiting were not identified, and it was concluded that the acute onset of vomiting was a consequence of the intussusception.

This case illustrated that PGI must be considered as another differential diagnosis for a gastric foreign body. In previous reports, the radiographic appearance of duodenogastric, pylorogastric, and gastrogastric intussusceptions were suggestive of the presence of a foreign body or a mass in the stomach, similar to the radiographic findings in this dog.^4–6,9,22 Due to this dog's radiographic findings, age, and acuteness of the vomiting, a gastric foreign body was considered the most likely diagnosis; thus, endoscopy and surgery were recommended. The diagnosis of PGI is exceedingly uncommon in dogs and was not initially considered. Abdominal ultrasonography would have been useful before and in place of endoscopy to differentiate PGI from a gastric foreign body and other causes of peracute vomiting, such as a small intestinal linear foreign body. In hindsight, an abdominal ultrasound might have eliminated the need for endoscopy and allowed earlier surgical intervention.

Spontaneous reduction of the intussusception in this case was unexpected, especially because attempts to reduce it endoscopically were unsuccessful. It has been suggested that spontaneous resolution of intestinal intussusceptions in dogs is more likely to occur in patients with a shorter duration of clinical signs.²³ In people, spontaneous reductions are more likely if they are confined to the small (versus large) bowel and the length of the intussusception is short.²⁴ Opioids, such as hydromorphone and fentanyl given to this dog preoperatively, would be expected to alter intestinal motility by inhibiting propulsive peristalsis while increasing nonpropulsive rhythmic contractions.²⁵ Opioids also affect gastric motility by slowing gastric emptying of liquids and solids.^25,26 This is associated with an elevated tonic contraction in the antrum and pylorus and decreased resting tone in the musculature of the gastric reservoir.²⁵ The prophylactic use of opioids has been suggested to reduce the rate of postoperative intestinal intussusception in dogs undergoing renal transplantation.²⁷ Therefore, it was feasible that the effects of preoperative opioids on gastrointestinal motility in this case might have contributed to the spontaneous resolution of the intussusception. Moreover, the distention of the stomach with air during endoscopy might also have been a contributing factor.²³ One of three previously reported dogs with PGI also had spontaneous resolution, but it was not stated whether this dog had received opioid analgesics.⁸

At surgery, the benefits and risks of performing a pyloropexy or duodenopexy were considered; however, these were not done because the intussusception was not present at the time of surgery. Duodenopexy was performed in one of the previous cases of PGI.⁴ It was unknown whether a duodenopexy was necessary, as recurrence of PGI was not reported.¹ The authors also speculated that duodenopexy might delay normal gastric emptying, which was another reason why this procedure was not performed.

Conclusion

This case report described a PGI in a young, large breed dog with acute onset of profuse vomiting and a mass-like lesion in the stomach on abdominal radiographs. Although PGI is a rare cause of acute vomiting, it should be considered when abdominal radiographs show a soft tissue mass effect in the stomach. Because PGI can be fatal, as shown in previous reports, further immediate diagnostic evaluation, such as ultrasonography, followed by emergency surgical intervention, is warranted. The authors acknowledge that an ultrasound is a less invasive alternative than endoscopy for documenting the presence of PGI. In hindsight, opioid administration might have been tried, which might have spared this dog from more invasive procedures, such as gastroscopy and laparotomy, if spontaneous resolution occurred. However, given the dog's severe presenting clinical signs, this would have been an unlikely decision. This case illustrated that the endoscopic appearance of a PGI was unique and diagnostic of PGI, thus providing an alternative minimally invasive means of differentiating PGI from other gastric mass lesions, such as foreign body, neoplasia, polyp, or granuloma. Unfortunately, the underlying cause of this PGI could not be determined in this case.