Introduction

Bilious vomiting syndrome (BVS) is described in veterinary textbooks as being associated with the early morning vomiting of bile in otherwise seemingly healthy dogs. It is thought to be due to the reflux of duodenal fluid into the gastric lumen and subsequent gastric mucosal irritation. The reflux may be associated with abnormal gastric motility due to the inability of the empty stomach to maintain intragastric pressure greater than duodenal pressure. The early morning timing of the vomiting may be the result of changes in the migrating motor complex of the small intestine seen during the overnight fasting state, but, as yet, no research has been done to test this hypothesis.¹ The presence of bile in the gastric lumen is thought to cause gastric mucosal damage.² BVS appears to be a diagnosis for patients with stereotypical clinical signs; however, it is unclear how many patients undergo a diagnostic workup sufficient to rule out specific causes of vomiting, such as parasites or inflammation. Although uncommon, intestinal obstruction or partial obstruction should also be ruled out.

The purpose of this retrospective study was to review cases of BVS in an effort to determine commonalities in signalment, diagnostic results, and treatments employed for dogs diagnosed with this condition. An additional aim was to identify cases where BVS was initially suspected but an alternate cause for vomiting was found.

Materials and Methods

Medical records from the Colorado State University Veterinary Teaching Hospital were searched for “canine” and “bilious vomiting syndrome” between the years of 2002 and 2012. Criteria for inclusion in this study were that the medical records were complete and that a diagnosis of BVS had been recorded at some point during the case history. Each medical record was reviewed, and the following information was retrieved: signalment, presenting complaint, duration of clinical signs before presentation, physical examination findings, diagnostic test results, treatments, and response to treatment.

Results

Twenty dogs were given the diagnosis of BVS during the study period. Twelve dog breeds were represented among the twenty canine patients included in the study. Those breeds included six mixed-breed dogs (30%); two each of Labrador retrievers, Scottish terriers, and Siberian huskies (10%); and one boxer, one Pomeranian, one West Highland white terrier, one Samoyed, one Yorkshire terrier, one Norwich terrier, one English bulldog, and one cocker spaniel (5%).

Of the 20 dogs, 10 (50%) were castrated males, seven (35%) were spayed females, and three (15%) were intact males. The ages of the affected patients ranged from 7 mo to 13 yr (median, 5.5 yr; range 7 mo–13 yr). Dogs weighed between 3 kg and 42 kg (median, 13.2 kg; range 3 kg–42 kg).

The client's presenting complaint for 13 of the 20 dogs (65%) was vomiting or chronic vomiting. Other presenting complaints included acid reflux, diarrhea, constipation, collapse, chronic intestinal problem, and “check liver.”

The duration of clinical signs was chronic (mo to yr) in 16 of the 20 dogs (80%). The median duration of clinical signs before presentation was 2.5 mo with a range of 3 days to 5.5 yr. The most commonly observed clinical signs included intermittent vomiting of bile (n = 17; 85%) and vomiting in the early morning (n=12; 60%). Other less commonly reported clinical signs included “lip smacking,” ptyalism, diarrhea, weight loss, polyuria/polydipsia, and decreased appetite (Table 1).

TABLE 1 Description of Clinical Signs, Diagnostics, and Treatments

View Fullscreen

Physical examination findings were unremarkable in 13 of the 20 dogs. Abnormalities encountered on physical examination included cranial abdominal pain (n=1; 5%), heart murmur (n=2; 10%), ptyalism (n=1; 5%), luxating patellae (n=1; 5%), otitis externa (n=1; 5%), and pyoderma (n=1; 5%). One dog in the study presented laterally recumbent with bile coming from the mouth and nose. This dog also had an increased respiratory rate and effort.

Diagnostics performed varied for each case, but consisted of complete blood cell count (n=15; 75%), biochemical profile (n=18; 90%), urinalysis (n=6; 30%), fecal floatation (n=3; 15%), fecal smear (n=2; 10%), abdominal radiographs (n=8; 40%), abdominal ultrasound (n=4; 20%), basal cortisol (n=1; 5%), adrenocorticotropic hormone (ACTH) stimulation test (n=2; 10%), bile acids (n=1; 5%), thoracic radiographs (n=1; 5%), and endoscopy with biopsies and subsequent histopathology (n=1; 5%).

Available laboratory data indicated mild clinicopathologic abnormalities in some dogs including eosinophilia, monocytosis, lymphopenia, basophilia, hyperglobulinemia, hypoalbuminemia, elevated alanine aminotransferase elevated alkaline phosphatase, and hyperglycemia. One dog had more significant laboratory abnormalities including elevated creatinine, calcium, globulins, albumin, cholesterol, bilirubin, creatine kinase, alkaline phosphatase aspartate aminotransferase gamma-glutamyl transferase and decreased chloride. Both dogs that underwent an ACTH stimulation test had a normal result. One dog had a basal cortisol of 1.08 ug/dL, but an ACTH stimulation test was not performed. The one dog where a serum bile acids test was performed had a normal result.

Abdominal radiographs were unremarkable in six of eight patients. Radiographs of one dog revealed a loss of serosal detail and hepatomegaly, and in a second dog they showed microhepatia and urolithiasis. Ultrasound revealed abnormalities in three of four dogs, including a hyperechoic liver (one dog), a suspect pancreatic mass (one dog), and an enlarged mesenteric lymph node (one dog). Thoracic radiographs were performed in one dog and revealed a bronchoalveolar pattern within the right lung fields, likely secondary to aspiration.

The one dog that underwent endoscopy of the upper gastrointestinal tract was found to have a dilated lower esophageal sphincter, no evidence of gross esophagitis, and no other abnormalities. Histopathology performed on gastric mucosa revealed evidence of scarring and fibrosis suggestive of a previous ulceration.

Multiple treatments were pursued with varying success. These treatments included adding a late evening meal (n=7; 35%), small frequent meals throughout the day (n=3; 15%), switching to a novel protein diet (n=4; 20%), deworming (pyrantel pamoate and/or fenbendazole) (n=9; 45%), addition of a gastric acid reducer (famotidine or omeprazole) (n=7; 35%), metoclopramide (n=6; 30%), cisapride (n=1; 5%), erythromycin (n=1; 5%), sucralfate (n=1; 5%), or switching to a prescription bland diet (n=2; 10%). Of the dogs that were dewormed, six were dewormed with a three day course of fenbendazole (50 mg/kg per os [PO] q 24 hr), 2 were dewormed with pyrantel (dose unknown), and 1 was dewormed with both.

Twelve dogs improved with therapy. The therapies that were most likely to produce an improvement in clinical signs, defined as a decrease in the frequency of vomiting or complete resolution of vomiting, included small frequent meals, a late evening meal, gastric acid reducers, and metoclopramide therapy. All three dogs that were given small frequent meals throughout the day had improvement of clinical signs (100%). Five of the seven dogs (71%) that were started on a late night meal or snack improved. Five of the seven dogs (71%) that were started on a gastric acid reducer also had decrease in frequency of vomiting or complete resolution. Four of the six dogs (67%) that were given metoclopramide improved. Five dogs did not improve or were lost to follow-up. The diagnosis of BVS was supplanted in three cases, where the eventual diagnosis was either gastric adenocarcinoma confirmed on necropsy, dietary indiscretion (ingestion of sand and plastic) and resultant gastroenteritis, or an unknown hepatopathy with vomiting of bile and resultant aspiration pneumonia. With the exception of the histopathology from necropsy to confirm a diagnosis of gastric adenocarcinoma, endoscopy was only performed in one other case where histopathology of the gastric mucosa revealed scarring and fibrosis suggestive of previous ulceration.

Discussion

In the current study, BVS was diagnosed most commonly in young, mixed-breed, castrated male dogs. The majority of these dogs responded to the treatments typically used for BVS.

Treatment options for BVS are generally focused on diet, gastric acid reduction, anti-emetic medication, and promotility agents. Some dogs respond to small, frequent feedings (i.e., greater than twice daily). A small amount of food later in the evening was also shown to be beneficial in the cases included in this study. One possible explanation for the effect of this intervention is that the food acts as a buffer to the refluxed bile or stimulates a normal post-prandial pattern of gastrointestinal motility.

Changing the diet to a prescription hypoallergenic diet, which includes both hydrolyzed protein and novel protein diets, also resulted in clinical improvement in some dogs. The specific type and composition of these diets was unknown. This is potentially important information, as the fat and fiber profiles may have impacted gastrointestinal motility and the response to dietary therapy. The positive response in some dogs is, however, consistent with an underlying food-responsive enteropathy or food intolerance.

Gastric acid reducers are frequently employed as a BVS treatment based on the assumption that gastric pH contributes the condition. Duodenal-gastric reflux is a component of a variety of human diseases often seen in children or after intestinal surgery. In humans, the term BVS is not used, but conditions with a similar pathophysiology are referred to as alkaline reflux gastritis or bile gastritis.³ Like BVS, these conditions are thought to result from some abnormalities in the motor function of the stomach and changes in the speed of gastric emptying.⁴ This results in reflux of bile into the stomach and subsequent gastric inflammation. Bile salts acting as detergents dissolve the mucosal lipids that help to form the gastric mucosal barrier, which allows for back diffusion of hydrogen ions and subsequent gastritis.^5,6 H2-receptor antagonists decrease acid secretion and may protect the mucosa from injury by back diffusion of acids through intracellular tight junctions damaged by exposure to bilious secretions. In humans with bile gastritis, most patients are hypo- or achlorohydric, and treatment with these agents is unrewarding.⁷ The two gastric acid reducers used in the cases included in this study were omeprazole and famotidine. Given the retrospective nature of this study, it is not known whether the dogs that received a gastric acid reducer and had improvement of their clinical signs were actually the result of the gastric acid reducer or concurrent therapies. It is also possible that the mechanism behind bilious vomiting in dogs may differ from humans. There are some H2-receptor antagonists that are also considered prokinetic, such as ranitidine and nizatidine, however the only H2-receptor antagonist employed in this study was famotidine. Omeprazole, a substituted benzimidazole proton pump inhibitor, and famotidine, a histamine H2-receptor antagonist, are frequently used to increase the gastric pH in both human and veterinary patients. A recent study compared the efficacy of oral famotidine and omeprazole on intragastric pH and found that omeprazole provided superior gastric acid suppression.⁸ However, the dogs in that study received a higher dose of these medications (famotidine 1.0–1.3 mg/kg PO q 12 hr; omeprazole 1.5–2.6 mg/kg PO q 24 hr) than what was prescribed to the dogs in this study (famotidine 0.5 mg/kg PO q 12 hr or 1 mg/kg PO q 24 hr; omeprazole 0.7–1 mg/kg PO q 24 hr). Once daily administration may prove to be more successful in treating dogs with BVS; however, controlled trials are needed to assess this.

Additional medications that were used for treatment of BVS included metoclopramide, cisapride, and erythromycin. Metoclopramide has been shown to be a good anti-nausea and anti-emetic agent, but its use as a prokinetic drug has been less successful.⁹ A single dose of metoclopramide before bedtime has been successful in some dogs with BVS.¹⁰ Also, administration of metoclopramide 30 min prior to feeding resulted in clinical improvement in some of the cases in this study. Cisapride was employed less frequently; however, it is a more potent prokinetic agent that may be beneficial in dogs with BVS.^11,12 Erythromycin is a motilin agonist that has been shown to improve gastric outflow, which may be helpful in reduction of duodenogastric reflux of bile leading to less gastric mucosal irritation.^13–15 Additional studies are needed to compare the efficacy of metoclopramide, cisapride, and erythromycin in treating dogs with BVS.

In human medicine, bilious vomiting is most commonly observed in neonates. Bilious emesis has been regarded as a hallmark for intestinal obstruction, and many textbooks and case reports emphasize that bile-stained vomit is considered a surgical emergency until proven otherwise.^16,17 In a previously published study, there were a similar proportion of neonates presenting with bile emesis that had no obvious surgical cause and were categorized as having idiopathic bilious vomiting.¹⁸ The etiology of nonsurgical bilious vomiting in infants is not completely known, but it is thought to result from a combination of gastroesophageal and duodenogastric reflux.¹⁹

In veterinary medicine, a diagnosis of BVS is usually given to dogs that are feeling well with the exception of intermittent vomiting of bile. BVS is considered a nonsurgical condition and is thought to be secondary to reflux of duodenal contents into the gastric lumen resulting in irritation of the mucosa.² Additional diagnostics aimed at assessing gastrointestinal motility, pH, and pressure within the gastric lumen may prove to be beneficial. Further evaluation of dogs diagnosed with BVS using a wireless motility capsule^a or indwelling simultaneous measurement of gastric and esophageal pH may prove informative. The wireless motility capsule system is a noninvasive method for assessment of gastric and intestinal motility that has been approved for use in humans and validated for use in dogs.^20–22

There were several limitations to the current study with the most important being its retrospective nature, small sample size, and lack of extensive diagnostic workup in many of the dogs. In a number of cases a more extensive diagnostic workup was not performed because the dog responded to treatment trials, so it was not determined if other underlying causes of vomiting had been excluded. Some cases lacked follow-up information to determine whether the dog continued to respond to the treatments employed, and some dogs were started on multiple concurrent treatments. Despite the above limitations, this is the first study that attempts to characterize the population of client-owned dogs that were given a diagnosis of BVS. This syndrome continues to be a clinical diagnosis, and clinical signs and response to therapy do not make it definitive. There are a number of other conditions that might respond to the treatments commonly employed for BVS, which could be misleading. BVS should be considered in any dog with intermittent vomiting of bile, especially if the vomiting occurs in the early morning, but it must continue to be considered a diagnosis of exclusion.

Conclusion

The canine patient most likely to be given a diagnosis of BVS was a young, mixed-breed, castrated male dog with a history of chronic vomiting of bile. Although the most frequently identified clinical sign in this population of dogs was the early morning vomiting of bile, 40% of the dogs given this diagnosis did not present with this clinical sign and, therefore, did not fit the textbook definition of the syndrome. This study also revealed that the diagnosis of BVS is being applied to patients in cases where a number of possible causes of the clinical signs have not been adequately ruled out. The therapies that most commonly produced a positive clinical response in these dogs were metoclopramide, gastric acid reducers, and late evening meals. Although this record of a clinical response may be consistent with several potential underlying causes, a much more extensive diagnostic workup would be required to identify specific therapeutic targets. The fact that several dogs given a diagnosis of BVS ended up having other very specific and serious conditions should motivate clinicians to make BVS truly a diagnosis of exclusion.