Esophago-arterial fistula secondary to bone ingestion has been previously reported only in association with the thoracic aorta. In this rare case, a chicken bone gradually eroded through the esophageal wall into the innominate artery causing esophago-innominate artery fistula. To our knowledge, this is the first such case reported. There are over 100 cases of esophagoaortic fistula secondary to swallowed objects, with only six reported survivors [1] [2] [3] [4] [5] [6]. This case illustrates the difficulty in making a diagnosis of this type, the hazards of delayed operative intervention, and some useful techniques in the management of a fistula.
A 77-year-old woman was referred to the gastroenterology department for evaluation of a suspected chicken bone lodged in her esophagus. Nine days previously, she had eaten chicken and felt a jabbing sensation in her lower neck and was unable to continue eating. From that point on she was able to swallow only liquids, as solids and semi-solids would be regurgitated shortly after ingestion. She went to a local emergency room the first day of her illness, where x-rays of the neck and chest failed to reveal any foreign body. She was told to wait a few days to see if the sensation would pass. One week later, she saw her regular physician due to the continued symptoms and weight loss. He ordered an upper GI x-ray, which demonstrated what appeared to be a small bone at the C6-C7 level lodged transversely in the esophagus. The next morning, she was evaluated in the gastroenterology department, and an esophagogastroduodenoscopy was performed. Although there had been no episodes of hematemesis, a fresh area of bleeding was noted at 19cm from the incisors and a large amount of blood was in the stomach. No foreign body was identified, but on initial passage of the endoscope there was difficulty moving past the area of blood in the upper esophagus, accompanied by some increased bleeding. The patient had a hypotensive episode immediately after the procedure, which seemed to respond to repeated doses of Narcan and was admitted for observation. Upon admission, her hemoglobin was 9.3. She was afebrile with an 11,700 WBC with normal differential. The PT was 11.7, PTT 20, and platelet count was 368,000.
![Esophago-innominate fistula from chicken bone – a case report figure 1](http://aws.labome.com/figure/te-905-1.png)
Approximately 12 hours after admission, the patient had a bout of hematemesis, passed melanotic stool, and became hypotensive. Her hemoglobin and hematocrit at the time were 5.9 and 18.1, respectively. She was transfused with crystalloid and packed red blood cells and was stable for six hours. Later, she again became hypotensive and had massive continued arterial hematemesis. At this point, an esophago-arterial fistula was suspected and she was rushed to the operating room for exploration.
Initially, the left neck was explored and no site of fistulization was identified. An esophagotomy was performed and the arterial hemorrhage was seen to be coming from the distal portion of the esophagus. A finger was inserted into the esophagotomy and a small defect could be felt distally and to the right. When digital pressure was applied to this defect, the hemorrhage stopped. Continued digital pressure was applied on the fistula from within the esophagus; median sternotomy and mediastinal exploration were performed. The communication between esophagus and artery was found at the posterior aspect of the innominate artery near its bifurcation into the right common carotid and right subclavian arteries. Due to the awkward position of the innominate artery defect, it was repaired through an anterior arteriotomy. No shunt was used, as adequate backpressure was felt to be present in the common carotid artery upon clamping. The esophageal wall at the fistula was black and necrotic. This was debrided and closed with double layer 2-0 chromic suture and a nasogastric tube was placed. The esophagotomy in the neck was closed in a similar fashion. Mediastinal and cervical drains were placed.
Postoperatively, the patient had a very stormy course. She suffered a stroke in the distribution of the right internal carotid artery, which was noted immediately postoperatively. She also had a consumptive coagulopathy, sepsis, pneumonia with respiratory failure requiring tracheostomy, and transient renal insufficiency. She was eventually transferred to a nursing home requiring tube feedings but no ventilator support.
Previous fistulization between the esophagus and aorta due to foreign body ingestion has been reported. Although we believe this to be the first report of fistulization between the esophagus and innominate artery due to foreign body, the signs, symptoms and course of this patient’s illness are almost identical to those of esophago-aortic fistula.
The first case of esophago-aortic fistula was reported by Dubrueil in 1818 [7]. In 1914, Chiari’s triad was described as [8] : 1) mid-thoracic pain, 2) sentinel arterial hemorrhage, and 3) final massive exsanguinating hematemesis after a symptom-free interval. The time between sentinel hemorrhage and exsanguination can range from minutes to years [9]. Esophago-aortic fistulas are most commonly caused by aortic aneurysms eroding into the esophagus, followed by esophageal or bronchial neoplasms. There have been cases where esophageal-vascular fistulas from nasogastric tube placements occur [10]. These have frequently been associated with aortic arch anomalies. An aberrant right subclavian artery has also been reported to cause a denovo right subclavian artery-esophageal fistula [11]. Foreign body impaction and other causes esophago-aortic fistulas are less frequent [12]. Foreign bodies in the esophagus tend to lodge in one of four narrow areas: 1) at or just below the cricopharyngeal muscle (as in this case), 2) at the aortic arch, usually 25-30cm from the incisors, 3) at the left mainstem bronchus, or 4) at the diaphragmatic hiatus. People with full dentures appear to be prone to foreign body impactions, especially bones, due to the decreased tactile discrimination while chewing. Our patient did in fact have full dentures.
Swallowed bones may be difficult to see on routine x-rays, especially in the mediastinal area. In one study, a chicken bone, which perforated the esophagus in the upper chest, could be seen in only 4 out of 54 x-rays taken [13]. Gastrografin or barium swallow may help delineate foreign bodies better, as seen in this case. This particular patient had no signs of perforation such as fever, subcutaneous or mediastinal emphysema, tachypnea, cyanosis, or abdominal tenderness. However, gastrografin swallow, possibly followed by barium swallow is usually diagnostic in a case of perforation. Although this patient did have a perforation, it was directly into an adjacent artery and the fistula was apparently plugged at the time of the barium study.
Perforations of the esophagus can occur immediately after a sharp object becomes lodged, but most often they occur at least 24 hours after the impaction due to progressive pressure necrosis of the esophageal wall and gradual migration of the object [14]. Perforations can also be caused by attempted endoscopic removal of the foreign body. It is generally agreed that a foreign body lodged in the esophagus should be removed as soon as the diagnosis is made because: "1) Once an object is impacted in the esophagus, the chance of spontaneous passage is small, 2) Edema from local trauma tends to grip the object more firmly making later manipulation increasingly difficult, and 3) Perforation of the esophagus is much more serious and dangerous than perforation of any other part of the gastrointestinal tract." [14] Signs or symptoms of esophageal foreign body, even if radiologic investigation is negative, warrant early esophagoscopy. Obviously, this was not done in our patient’s case, and the esophageal foreign body, most likely a chicken bone, gradually eroded into the nearby vessel.
Our patient’s first herald bleed could be considered to have occurred at the time of her esophagoscopy, due to the blood noted at 19cm and in the stomach. Although there was no prior hematemesis or melena, this endoscopic finding along with her history was enough to warrant surgical exploration at that time to search for esophago-arterial fistula. A sentinel hemorrhage is believed to be the indication for urgent operation when esophago-aortic fistula is suspected [1] [2] [12] [15]. In actuality, the diagnosis of such a fistula was not entertained until late, when the massive bleed occurred. At that time, a Sengstaken-Blakemore tube could have been inserted for a temporizing measure. This was used successfully by McFadden and Dang for overnight control of bleeding from an esophago-aortic fistula at a balloon pressure of 80mmHg with no mucosal damage [16]. In another case, arteriographic tamponade of an esophago-aortic fistula was accomplished with bucrylate [17]. At the time of surgery on the reported patient, the arterial defect was repaired without graft or patch. However, there have been some late hemorrhages from simple aortic repairs in areas of contamination from esophageal perforation. Some authors recommend resections of the contaminated portion of vessel and replacement with graft [12] [16]. Late hemorrhage did not occur in this case and the esophageal and arterial repairs should be separated by interposed tissue to try to prevent recurrent fistulization [9].
- Carter R, Mulder G, Snyder E, Brewer L. Aortoesophageal fistula. Am J Surg. 1978;136:26-30 pubmed
- Ctercteko G, Mok C. Aorta-esophageal fistula induced by a foreign body: the first recorded survival. J Thorac Cardiovasc Surg. 1980;80:233-5 pubmed
- Yonago R, Iben A, Mark J. Aortic bypass in the management of aortoesophageal fistula. Ann Thorac Surg. 1969;7:235-7 pubmed
- Podbielski F, Rodriguez H, Zhu R, Worley T, Fontaine J, Connolly M. Aortoesophageal fistula secondary to reflux esophagitis. Dig Surg. 2007;24:66-7 pubmed
- Montgomery A, Chilvers A. Oesophago-aortic fistula. Postgrad Med J. 1981;57:380 pubmed
- Snyder D, Crawford E. Successful treatment of primary aorta-esophageal fistula resulting from aortic aneurysm. J Thorac Cardiovasc Surg. 1983;85:457-63 pubmed
- Dubrueil: Observation sur la perforation de l’oesophage et de l’aorte thoracique par une potion d’os avale: avec des reflexions. J Univ Sci Med 1818; 9:357.
- Chiari H: Ueber premdkorperverletzung des oesophagus mit aortenperforation. Ber Klin Wochenschr. 1914; 51:7.
- Henry W, MISCALL L. Aortic-esophageal fistula. J Thorac Cardiovasc Surg. 1960;39:258-62 pubmed
- Minyard A, Smith D. Arterial-esophageal fistulae in patients requiring nasogastric esophageal intubation. Am J Forensic Med Pathol. 2000;21:74-8 pubmed
- Millar A, Rostom A, Rasuli P, Saloojee N. Upper gastrointestinal bleeding secondary to an aberrant right subclavian artery-esophageal fistula: a case report and review of the literature. Can J Gastroenterol. 2007;21:389-92 pubmed
- Wilson R, Dean P, Lewis M. Aortoesophageal fistula due to a foreign body. Gastrointest Endosc. 1987;33:448-50 pubmed
- So S, Mok C, Lam W, Yu D. Haemoptysis due to unsuspected foreign body penetration of the oesophagus. Aust N Z J Med. 1982;12:533-5 pubmed
- Nandi P, Ong G. Foreign body in the oesophagus: review of 2394 cases. Br J Surg. 1978;65:5-9 pubmed
- Patel N, Sangchantr W. Aortoesophageal fistula and sentinel hemorrhage. IMJ Ill Med J. 1985;167:58-9 pubmed
- McFaddin D, Dang C. Management of aortoesophageal fistula. A case report. Am Surg. 1985;51:548-50 pubmed
- Reedy F. Embolization of aortoesophageal fistula: a new therapeutic approach. J Vasc Surg. 1988;8:349-50 pubmed