Congenital oesophageal stricture in a Japanese shiba inu

Introduction

Congenital oesophageal stricture is defined as intrinsic stenosis of the oesophagus caused by congenital malformations of the oesophageal wall architecture (Amae and others 2003). The oesophagus is formed from the portion of foregut between the caudal limit of the pharynx and the fusiform enlargement of the stomach. Initially very short, this portion elongates as the heart descends from the neck into the thorax. The oesophagus is also involved in the origin of the lower respiratory tract because the foregut endoderm supplies the lining epithelium of the respiratory organs as they form from the mesoderm. At one stage, the rapid proliferation of the endodermal lining completely obstructs the lumen, but the passage is later restored (Dyce and others 1996).

Congenital oesophageal stricture in humans is rare: the incidence is estimated at 1:25,000 to 1:50,000 live births (Bluestone and others 1969), and approximately 500 cases have been reported to date. Associated congenital anomalies reported in one-third of these cases include oesophageal atresia, intestinal atresia, cardiac defects and chromosomal anomalies (Fekete and others 1987).

In humans, three types of congenital oesophageal stricture are described (Fekete and others 1987): fibromuscular thickening, which is associated with reduced myenteric nitrinergic neurons (Singaram and others 1995); cartilaginous ring, which results from sequestration of tracheobronchial precursor cells in the oesophageal wall caused by defective embryologic separation of the primitive foregut from the respiratory tract (Anderson and others 1973, Rose and others 1975); and membranous web, which results from a disturbance of recanalisation (Amae and others 2003). These lesions usually arise alone, but in rare cases, they can arise simultaneously (Takamizawa and others 2002, Nagae and others 2005).

Congenital oesophageal stricture is a rare abnormality in animals. There are reports of two affected foals (Stewart and Reinertson 1983, Clabough and others 1991) and two affected dogs (Schnelle 1931, Pande and others 1995). Neither of the latter reports ascertained the underlying pathological changes.

Case history

An 11-week-old, female Japanese shiba inu with a history of intermittent regurgitation of both solids and liquids since birth was presented to the Queen Mother Hospital for Animals. The dog had been weaned on semisolid food (Hill’s a/d) at seven weeks of age and was fed from an elevated position. Physical examination revealed a bright, small puppy in poor body condition, weighing 1·3 kg, but was otherwise unremarkable. Haematologically there was reduced mean red blood cell volume (59·4 fl: reference range 60 to 77 fl) and reduced mean red blood cell haemoglobin concentration (17·7 pg: reference range 19·5 to 24·5pg) compatible with nutritional iron deficiency.

An oesophagram using barium sulphate suspension (Baritop 100; Sano Chemia Diagnostics UK Ltd) mixed with Hill’s a/d diet demonstrated a persistent, focal, curvilinear filling defect within the caudal cervical oesophagus and dilatation of the oesophagus compatible with partial obstruction (Fig 1). Some food passed this lesion and entered the stomach. The caudal oesophagus had a normal appearance and motility.

Further investigations were under general anaesthesia. Following premedication with 4 mg/kg pethidine (Pethidine; Arnolds) and 0·3 mg/kg midazolam (Hypnovel; Roche) intramuscularly, anaesthesia was induced with sevoflurane (Sevo-Flo; Abbott Animal Health) and maintained with isoflurane (Isoba; Schering Plough Animal Health) in oxygen.

Oesophagoscopy showed normal oesophageal mucosa and confirmed the presence of a non-distensible band of tissue within the lumen. Transthoracic ultrasound did not reveal any abnormalities. Non-selective angiography with two 2 ml boluses of sodium meglumine diatrizoate (Urografin 370; Schering) injected through a 16 gauge jugular catheter revealed normal anatomy of the aorta and its branches (Fig 2a,b).

Mechanical ventilation was instituted, and neuromuscular blockade was achieved with 0·05 mg/kg cisatracurium intravenously (Nimbex; Abbott Laboratories). The affected part of the oesophagus was exposed through a ventral midline cervical approach and a partial, rostral median sternotomy. Passage of an oro-oesophageal tube until resistance was encountered allowed intraoperative identification of the stricture. The external surface of the oesophagus and adjacent blood vessels appeared normal. The oesophagus oral to the stricture was dilated. A 4 cm longitudinal incision centred at the level of the obstruction was made into the oesophagus, and a membranous band was visualised occluding the lumen (Fig 3). This was resected following submucosal dissection, and the oesophagotomy closed in a transverse manner in a single layer with a simple interrupted pattern of 2 metric polydioxanone (PDS II; Ethicon). The thorax was closed with 0 metric polydioxanone sutures placed in a figure-of-eight pattern to appose the sternebrae. The pectoral muscles were sutured with 2 metric polydioxanone, and then the subcutaneous tissues and skin were closed routinely. Perioperative antibiotics comprised one dose of 20 mg/kg cefuroxime (Zinacef; Glaxo) intravenously. Perioperative analgesia comprised 4 mg/kg pethidine intramuscularly and 2 mg/kg carprofen (Rimadyl; Pfizer) intravenously. Immediate postoperative radiographs documented bilateral pneumothorax, which was drained by needle thoracocentesis before recovery.

Morphine (Morphine Sulphate; Martindale Pharmaceuticals) administration at 0·3 mg/kg intravenously four hours after surgery caused respiratory arrest; intermittent positive pressure ventilation was initiated, and 0·02 mg/kg naloxone (Narcan; Myers Squibb) was given twice intravenously. Normal breathing resumed, and further opiate use was discontinued.

Oral feeding with small volumes of Hill’s a/d was continued postoperatively, and there was no regurgitation. A barium oesophagram was performed 72 hours postoperatively. This showed persistence of mild dilatation of the cervical oesophagus, no stricture (Fig 4) and subjectively normal oesophageal motility caudal to the thoracic inlet. The dog was discharged with instructions for postural feeding to be continued and for a commercial puppy food to be fed.

Histological examination of the resected tissue showed focally hyperplastic epithelium and dense fibrous underlying connective tissue with no inflammation (Fig 5).

At follow-up two years postoperatively, the dog had grown to the breed’s standard size, had no other abnormalities and was able to eat softened biscuits. Regurgitation was noted rarely when the dog stole food and ate rapidly.

Discussion

In a young animal with regurgitation since birth, differential diagnoses include the following: anatomical causes (vascular ring anomaly, cricopharyngeal dysfunction, hiatal hernia, oesophageal diverticulum and broncho-oesophageal fistula); neuromuscular causes (congenital megaoesophagus); obstructive oesophageal causes (foreign body, retropharyngeal lymphadenopathy, mediastinal mass, cervical mass, stricture and neoplasia); oesophagitis (traumatic, reflux, irritant or secondary to, for example, megaoesophagus) and anomalous causes (gastro-oesophageal intussusception) (Elwood 2006). When such cases are presented for investigation, initial diagnostic efforts should concentrate on ruling out treatable primary disease and identifying complications such as aspiration pneumonia. A similar diagnostic plan is adopted in most cases.

Plain thoracic radiographs provide an initial survey and along with clinical signs can identify aspiration pneumonia. Oesophagography enables assessment of motility and swallowing function as well as identification of the number, location and nature of any obstructive lesions. Oesophagoscopy may only enable visualisation of the cranial aspect of the most cranial lesion but allows mucosal inspection and assessment of mural integrity. Angiography is useful to identify and characterise vascular ring anomalies (House and others 2005). In this case, oesophagography provided the clearest indication of the lesion, and other diagnostic tests ruled out other lesions.

Treatment options for oesophageal stricture are dilatation, via bougienage or balloon dilatation, or surgical resection. While dilatation techniques are minimally invasive, they carry the risks of oesophageal perforation and worsening stricture caused by gastro-oesophageal reflux because of repeated anaesthesia. The morbidity of oesophagotomy must be weighed against the benefits of direct visualisation for precise and complete surgical resection as well as the recovery of stricture tissue for histological analysis. Oesophagotomy can allow simple excision of the stricture tissue as in the case reported here, but full-thickness resection and anastomosis may be required in some cases.

In human surgery, surgical resection is considered to be more effective than dilatation techniques (Amae and others 2003, Zhao and others 2004) but may not be curative if postresection stricture develops (Vasudevan and others 2002). Selection of treatment may be based on histological type, such that tracheobronchial remnants require surgical resection while fibromuscular thickening and membranous webbing can resolve with dilatation (Vasudevan and others 2002). However, it is difficult to anticipate the histology of the lesion before surgical excision.

In the case reported here, the decision to undertake surgical exploration was made because non-invasive diagnostic tests had not yielded a definitive diagnosis. Surgical exploration and resection of the abnormal tissue allowed both diagnosis and management. In this case, the surgical appearance and histological examination suggest that the stricture corresponds to the membranous web form described in human beings, and surgical treatment resulted in excellent long-term outcome.