Phenobarbital-Responsive Ptyalism, Dysphagia, and Apparent Esophageal Spasm in a German Shepherd Puppy

Case Report

A 10-week-old, male German shepherd dog was presented to the University of Wisconsin Veterinary Medical Teaching Hospital with a primary complaint of periodic ptyalism, difficulty swallowing, and vomiting for at least 2 weeks prior to presentation. The new owner reported that episodes would occur four to six times per day, with no apparent stimulus. During these episodes, the puppy appeared to be extremely distressed, behaved as though he were in pain, and seemed to have trouble breathing. There was no evidence of muscle weakness, and the animal’s mentation was normal. The episodes lasted approximately 15 minutes, after which the puppy would sleep. The animal would often vomit during or soon after the attacks, but he had a ravenous appetite between episodes. No diarrhea or abnormal bowel movements were noted. The dog had received a combination distemper-adenovirus-parainfluenzapar-vovirus vaccine, a rabies vaccine, and an unspecified anthelmintic prior to presentation. The puppy was significantly smaller than his littermates, and no similar episodes were noted in the other littermates. Thoracic radiographs and an esophagram performed prior to referral were unremarkable.

On presentation, the puppy was bright, alert, and responsive, but had just vomited a large amount of bile-stained liquid. Temperature, pulse, and respirations were within reference ranges. The puppy was thin (10.8 kg), with a body condition score of 1/5.¹ The capillary refill time was <2 seconds, mucous membranes were pink but tacky, and dehydration was estimated at 7%. There was no evidence of halitosis, lingual lesions, or dental disease on oral examination. Cranial nerve (CN) reflexes were normal, and there was no evidence of muscle atrophy of the head or neck. The remainder of the neurological examination was normal. The mandibular salivary glands were moderately enlarged but were not firm or painful on palpation. Auscultation of the chest did not reveal any abnormalities; however, the puppy had a spontaneous soft cough in the examination room. Gas-distended bowel loops and moderate discomfort were found upon abdominal palpation.

A complete blood count (CBC) showed mild neutropenia (2.53 × 10³ cells/μL; reference range, 3.0 to 11.5 × 10³ cells/μL) and eosinophilia (0.87 × 10³ cells/μL; reference range, 0.1 to 0.75 × 10³ cells/μL). A serum biochemical profile revealed mild hypercalcemia (12.4 mg/dL; reference range, 9.5 to 11.2 mg/dL) and hyperphosphatemia (8.1 mg/dL; reference range, 2.6 to 6.2 mg/dL) consistent with growth. Thoracic radiographs showed a fluid-filled esophagus, microcardia consistent with hypovolemia; no pulmonary infiltrates were seen. Abdominal radiographs demonstrated a dilated, gas-filled stomach and small intestines.

The puppy was admitted for observation and intravenous fluid therapy for the apparent dehydration. The puppy vomited once during the evening. The next day an esophagram with fluoroscopy and an upper gastrointestinal (GI) barium series were performed. There was no evidence of pharyngeal or esophageal dysfunction on the esophagram. The upper GI series showed normal motility with irregular margins in some areas of the small intestine, suggestive of enteritis. Because of the puppy’s breed, ravenous appetite, and thin body condition, assays for serum trypsinlike immunoreactivity, folate, and cobalamin levels were run to rule out exocrine pancreatic insufficiency, and all results were within reference ranges. A fecal sample was submitted for a direct smear examination and fecal flotation, both of which were negative for parasites.

During hospitalization, several episodes of ptyalism and repeated attempts at swallowing were observed. There was no evidence of cyanosis or tachypnea during the episodes. Increased upper airway sounds and a soft cough were occasionally noted and were attributed to excessive saliva in the oropharynx. Differential diagnoses for the ptyalism included difficulty swallowing from pharyngeal or esophageal dysfunction or palsy of one or more cranial nerves, and excessive production of saliva from nausea, gastric ulcers, sialoadenitis, stomatitis, pharyngitis, esophagitis, excess parasympathetic stimulation, etc. Given the normal CN reflexes and normal pharyngeal and esophageal function on fluoroscopy, both cranial nerve dysfunction and pharyngeal/esophageal dysfunction were considered unlikely. The puppy did not appear to be depressed or nauseous just prior to the episodes, and there was no anemia, hypoproteinemia, melena, or abdominal pain to suggest the presence of gastric ulceration. Although ptyalism may occur with hepatic encephalopathy and portosystemic shunts in cats, this puppy had no signs of altered mentation during the episodes, and the abrupt onset and cessation of signs were not consistent with a metabolic encephalopathy. In addition, no evidence of a portosystemic shunt was found in the CBC (e.g., microcytosis or target cells) or serum biochemical profile (e.g., alterations in alanine aminotransferase, albumin, blood urea nitrogen, cholesterol, glucose).

Empirical therapy for esophagitis was started with famotidine (1 mg/kg per os [PO] q 24 hours), sucralfate (500-mg slurry PO q 6 hours), and metoclopramide (0.2 mg/kg PO q 8 hours). Because of the eosinophilia, poor body condition, vomiting, and suggestion of enteritis on the barium series, the puppy was also treated with fenbendazole (50 mg/kg PO q 24 hours for 3 days). Although eosinophilia can be a marker for hypoadrenocorticism in dogs, particularly when accompanied by hypovolemia, hypoadrenocorticism was considered unlikely in this very young dog given the enlarged salivary glands, the abrupt onset and cessation of the signs, and the lack of azotemia or electrolyte abnormalities on the biochemical profile.

To rule out possible salivary, oropharyngeal, or gastroesophageal disease, a thorough oral examination and endoscopy were performed under general anesthesia. The oral examination revealed no evidence of foreign bodies, swelling, obstruction, or ulceration. On palpation of the neck, a smooth, firm, thickened area in the midcervical region of the esophagus was detected. Upper GI endoscopy revealed a 2-cm narrowing in the esophagus, beginning approximately 10 cm from the commissure of the lips. The 8-mm endoscope could be passed easily through the narrowed region to the gastroesophageal sphincter, which was grossly normal. The stomach and duodenum were normal in appearance, and multiple mucosal biopsies were obtained from both sites. In addition, gastric tissue was submitted for urease testing for Helicobacter organisms.

Ultrasonography of the neck was performed to further evaluate the palpable thickening near the narrowed region of the midcervical esophagus. A stomach tube was placed in the esophagus to the level of the palpable thickening to enhance imaging. Ultrasonography revealed a focal, circumferential thickening of the muscular layer of the esophagus. There was no cervical lymphadenopathy, and the enlarged mandibular salivary glands were of normal echogenicity. No evidence of inflammation or necrosis was found on cytopathology of fine-needle aspirates of both mandibular salivary glands or the thickened area of the esophagus. The gastric mucosal urease test was negative. Histopathology of the stomach and intestinal biopsies were essentially normal, with only minimal mononuclear and eosinophilic infiltrates found in the duodenum.

The focal narrowing of the esophagus seen on endoscopy and the localized circumferential thickening of the muscular layer of the esophagus detected on ultrasonography were attributed to muscular spasm of the esophagus. Muscular spasm, rather than hypertrophy or infiltrative disease, seemed most likely given the apparent intermittent nature of the esophageal narrowing (the esophagram was normal 3 days previously, and the esophageal thickening was not palpable on the initial visit). Although a dynamic change in the esophagus was not observed at the time of endoscopy and ultrasonography, intermittent esophageal spasm was also supported by the dog’s intermittent episodes of difficulty swallowing and his normal ability to prehend and swallow solid food between those episodes.

Given the normal examination and histopathological findings and the pronounced episodic nature of the clinical signs, a neurological cause seemed likely. Since no parasympathomimetic drugs had been administered and no other signs of excess parasympathetic stimulation (such as bradycardia, miosis, urinary incontinence, or diarrhea) were present, a diagnosis of limbic epilepsy was considered.

A syndrome of phenobarbital-responsive hypersialosis with salivary gland enlargement has been reported in dogs and has been postulated to be a form of limbic epilepsy. Some affected dogs have abnormal electroencephalographic (EEG) tracings.² Equipment problems precluded obtaining an EEG during the dog’s hospitalization. While this relatively rare syndrome has not been known to be associated with esophageal spasm, it is characterized by episodic ptyalism, enlarged and sometimes painful salivary glands, and intermittent dysphagia, retching, or vomiting. A therapeutic trial of phenobarbital was instituted. The dog was given a loading dose of phenobarbital (12 mg/kg intramuscularly [IM]) and was discharged on oral phenobarbital (3 mg/kg q 12 hours).

Over the next 3 days, the owner reported that for the first time since acquiring the puppy at 8 weeks of age, the episodes of ptyalism were measurably decreased in both severity and frequency. At the end of 7 days of therapy, the puppy was normal. Two weeks after starting the phenobarbital, the puppy was reevaluated. He had gained 5.2 kg (body weight, 16 kg) and had visibly grown in stature. A CBC and hepatic values on a serum biochemical profile were within reference ranges, but the serum phenobarbital concentration (10 μg/mL) was below the therapeutic range (15 to 40 μg/mL). The dosage of phenobarbital was increased by 20%. The animal was reevaluated by the referring veterinarian 3 weeks later. The dog remained normal and was in excellent body condition. Serum phenobarbital concentration at that time was 19 μg/mL.

Over the next month, while the puppy was boarded at a kennel facility, the phenobarbital was inadvertently discontinued. Upon return, the owner reported a relapse of the ptyalism, dysphagia, and enlargement of the salivary glands. The animal was given a loading dose of phenobarbital, and oral-maintenance phenobarbital therapy was reinstituted as previously described. Once again, the signs of ptyalism and dysphagia resolved. The puppy remained free of clinical signs for 6 months while on phenobarbital treatment. At a 6-month recheck examination, the animal was in excellent body condition, and a CBC, serum biochemical profile, and bile acids were within reference ranges. The serum phenobarbital concentration was in the low therapeutic range at 17.2 μg/mL. Endoscopy of the esophagus was performed and was normal. There was no evidence of esophageal narrowing at any level of the esophagus. Ultrasonography of the esophagus was also performed, which showed complete resolution of the focal esophageal thickening.

After 10 months of treatment with phenobarbital, the animal was still free of clinical signs, and the owner agreed to a trial reduction of the phenobarbital dosage. The dosage of phenobarbital was decreased by 25% every 3 weeks until the drug was discontinued. The dog has been off of phenobarbital therapy for >6 months, with no relapse of clinical signs.

Discussion

Mandibular salivary gland enlargement, often accompanied by ptyalism, has been reported previously in 43 dogs.^2–10 The earliest reports were in small terriers and were characterized by salivary gland necrosis on biopsy [see Table; case nos. 1, 2, 3].³¹⁰ One of these dogs (case no. 1) had a normal EEG but nevertheless responded to phenobarbital. Two other dogs (case nos. 2, 3) treated only with sialoadenectomy had no response and were euthanized.

A subsequent series of case reports, primarily in terriers and spaniels, showed variable necrosis on either aspiration cytopathology or histopathology of the salivary glands.^4–7 One of these dogs (case no. 4) had a normal EEG. None of the dogs responded to sialoadenectomy, but all of them responded to phenobarbital.^4–7 A series of 19 dogs with mandibular salivary gland enlargement and pain, retching, or ptyalism were reported from South Africa [see Table; case nos. 9 through 27].⁸ Ten of the dogs had inflammation or necrosis of the salivary glands, while eight dogs had normal salivary cytopathology or histopathology.⁸ Interestingly, 14 of the dogs had underlying esophageal disease, such as esophagitis, megaesophagus, or granulomas from Spirocerca lupi. In some of these dogs, the mandibular salivary gland enlargement and pain resolved after the esophageal disease was treated, suggesting a possible causal relationship between esophageal disease and salivary gland enlargement. The mechanical relationship between esophageal disease and salivary gland pathology was not established; however, a neural reflex (i.e., vagal afferent input from the esophagus leading to secondary parasympathetic or neuroendocrine effects on salivary blood flow) was postulated.⁸

Recently three dogs were reported with ptyalism and mandibular salivary gland enlargement that were responsive to phenobarbital [see Table; case nos. 28, 29, 30].² Histopathology of the salivary glands was normal in all three dogs; an EEG performed in one dog was consistent with epileptiform discharges. Additionally, a series of 13 dogs with salivary gland enlargement and clinical signs of ptyalism, retching, and gulping were described in a study of sialadenosis in dogs [see Table; case nos. 31 through 43].⁹ Nine of these dogs had no histopathological abnormalities in their salivary tissue, but four had mild ductule hypertrophy. Despite a thorough evaluation, no underlying cause could be found for the sialadenosis. All 13 dogs had an apparent response to phenobarbital therapy, with resolution of salivary gland enlargement and clinical signs.

Salivary gland disease associated with esophageal disorders is not well recognized in humans. However, salivary gland enlargement (i.e., sialadenosis) resulting from dysfunction of the autonomic nervous system has been described in people.¹¹ Although histopathology of the salivary glands is often normal, electron microscopy has shown degenerative changes in postganglionic neurons (e.g., loss of granules, hydropic swelling), and associated epithelial cells may undergo necrosis.

It is unclear whether dogs with salivary gland enlargement and necrosis suffer from a different condition than do dogs with the same clinical signs but no salivary gland pathology. Fine-needle aspirates may miss focal areas of necrosis, but some affected dogs have had no evidence of salivary necrosis even at necropsy.⁸ It is possible that the necrosis observed was a result of the underlying disease process (such as autonomic dysfunction, as has been described in humans), rather than a direct cause of the clinical signs.⁸ This theory was supported by the fact that removal of the affected salivary glands in many of the dogs did not result in resolution of the clinical signs.⁸

The pathogenesis of this idiopathic phenobarbital-responsive hypersialosis (or sialadenosis) in dogs is not well understood. A form of limbic epilepsy has been proposed, based on the episodic nature of the clinical signs, lack of oropharyngeal or gastroesophageal disease in many dogs, and EEG tracings consistent with seizure activity.² However, dogs with normal EEG studies have also responded to phenobarbital therapy.⁴⁸ In the puppy of this report, inadvertent discontinuation of phenobarbital led to a relapse of ptyalism and dysphagia, supporting the responsiveness of this syndrome to phenobarbital.

Although the response of this syndrome to anticonvulsants suggests that it may be a form of epilepsy, it is also possible that phenobarbital acts via alternative mechanisms. Even though there is no evidence in the literature that phenobarbital has a direct effect on saliva production or esophageal motility, this drug does inhibit small intestinal motility and could conceivably have local rather than central effects on this disorder.¹² In addition, esophageal spasm in humans can be precipitated by stress, and it is therefore possible that phenobarbital was acting as a nonspecific sedative in this dog.¹³

This case is unique because of the early age of onset (≤8 weeks) versus the previously reported cases (6 months to 12 years). In addition, this is the first report of presumed esophageal spasm associated with these episodes. The esophageal narrowing documented by endoscopy and ultrasonography also appeared to be episodic. The question remains as to whether the ptyalism of these affected animals develops from an intermittent inability to swallow saliva or from hypersalivation. Diffuse esophageal spasm has been reported rarely in humans and may be associated with dysphagia and chest pain.¹⁴ Abnormal esophageal contractions in people may follow cholinergic stimulation; however, ptyalism is not a feature of this condition. The vomiting exhibited by this animal and other previously reported cases could have been secondary to salivary and/or esophageal pain, or it may have been a separate central manifestation of limbic epilepsy.¹⁵

Conclusion

Although rare, the syndrome of phenobarbital-responsive hypersialosis should be considered in dogs with episodic ptyalism, enlarged salivary glands, and normal findings on cranial nerve testing and examination of the salivary glands, oropharynx, esophagus, and stomach. The findings in the case reported here suggested that intermittent esophageal spasm may be a feature of this syndrome and may have contributed to the ptyalism and dysphagia associated with this case. Treatment with phenobarbital was successful in resolving the clinical signs in this dog, and it was discontinued after several months, without relapse. The pathogenesis of this syndrome is not yet understood, but it may represent a form of limbic epilepsy or peripheral autonomic dysfunction.