Diagnosis and therapy of esophageal perforation in the aged

Abstract

AIM: To study the diagnosis and therapy of esophageal perforation in the aged.

METHODS: Three-hundred-and-thirty-five elderly patients undergoing endoscopy or therapeutic esophageal dilation in our hospital between July 1988 and August 1995 were studied retrospectively. Of these patients, 31 had esophageal perforation, 17 of whom were treated nonsurgically (nasogastric drainage, antibiotics, and intravenous alimentation), 8 underwent total esophagectomy, and 6 received surgical drainage.

RESULTS: Sixteen of 31 patients with esophageal perforation were cured. Of the patients in whom the perforation healed, 9 had persistent dysphagia while swallowing solid foods. Six patients died. The cure rate was 51.6% and the mortality rate was 19.4%.

CONCLUSION: Aged people are subject to esophageal perforation. The main typical symptoms of esophageal perforation are chest pain, shortness of breath and dysphagia; It is not difficult to make a correct diagnosis, but early diagnosis and treatment are crucial to the saving the patient’s life.

Key Words: Esophageal perforation/etiology, Esophageal perforation/surgery, Esophageal perforation/diagnosis, Esophageal perforation/therapy

INTRODUCTION

Esophageal perforation is a serious complication of instrumental examination of the esophagus and also occurs with increased luminal pressure (barogenic perforation) that usually results from vomiting (Boerhaave′s syndrome)[1]. With advances in antibiotic therapy and a heightened awareness of this complication, patients receiving early treatment may survive and many of them can be cured.

MATERIALS AND METHODS

Patients

Thirty-one patients (23 men and 8 women; mean age: 68±4 years, range: 60-79 years) underwent endoscopy or therapeutic esophageal dilation in our hospital between July 1988 and August 1995.

Etiology

In 24 patients, esophageal perforation occurred after instrumental examination, including 14 patients undergoing dilation for benign esophageal stricture (13 chronic strictures caused by reflux esophagitis and 1 radiative stricture). In the 14 patients, the perforation occurred in 6 after dilation with Savary dilators performed under fluoroscopic guidance, and in 6 after dilation with Maloney dilators that was completed without fluoroscopy. Among the latter 6 cases, 1 perforation occurred and enlarged after endoscopic balloon dilation for benign esophageal stricture, 1 occurred after pneumatic dilation for achalasia, 1 occurred as a complication of sclerotherapy, and 3 were caused by attempts of intubation (1 by a rigid esophagoscope, 1 by endotracheal intubation, and 1 by endoscopic retrograde cholangiopancreatography with a flexible endoscope for a Zenker′s diverticulum). The last one was the only case of perforation by diagnostic flexible endoscope during this period. Perforation occurred in 6 patients after emergency gastroscopy for stomachache, in 5 after vomiting (barogenic perforation), in 1 after blunt trauma to the chest and in 1 without any definite causes (spontaneous perforation).

Symptoms

The main clinical symptoms in esophageal perforation were chest pain, dysphagia (unable to take food via mouth) and shortness of breath. Chest pain was the most frequent symptom, and was noted in 28 of the 31 cases (90.3%). Twelve patients (38.7%) had dysphagia and 6 (19.4%) complained of shortness of breath.

Diagnosis

Chest radiography was performed in 22 of the 31 patients and revealed mediastinal emphysema in 2; twenty cases were diagnosed as perforation of the esophagus by contrast study. Results of barium swallow studies were initially normal in 2 cases but a repeated examination several hours later showed the esophageal perforation. Another patient with a normal barium swallow showed mediastinal emphysema on the chest radiograph and did not undergo the contrast study. Three patients were diagnosed as having perforation of the esophagus by gastroscopy, and 1 perforation was recognized during the examination with rigid endoscopy. Of the 31 perforations, 21 occurred in the upper esophagus, 6 in the midesophagus, and 4 in the lower esophagus.

Criteria for treatment of perforation

The choice of therapy for perforation was based on clinical criteria. Patients were treated nonsurgically if they met the following criteria described by Cameron et al[2]: recently-developed perforation (within 24 h); no food intake after the perforation; perforation not proximal to a high-grade stenosis; minor symptoms (pain or dysphagia) without clinical signs of sepsis or hemodynamic compromise; perforation contained within the mediastinum without contamination of adjacent body cavities (e.g. pleural space); contrast studies showing a small perforation with good drainage of contrast material back into the esophagus.

Esophagectomy should be performed if longstanding suppuration exists at the periphery of the perforation or underlying diseases (e.g. mega-esophagus, neoplasm and chronic refractory stricture) lead to dysfunction of the esophagus, or even to closure of perforation. The operation must be decided based on the general condition of the patient. The diagnosis of esophageal perforation was established within 24 h in 25 patients. All of the patients with suspected perforation were initially treated with nasogastric aspiration and intravenous antibiotics. Seventeen patients were treated nonsurgically (nasogastric drainage, antibiotics, and intravenous alimentation) using the criteria mentioned above. Six patients underwent surgical drainage, and 8 patients underwent total esophagectomy (Table 1). Serious contaminations of the pleural space and mediastinitis were noted in 1 patient. Because of purulent mediastinitis, the exact location of the perforation could not be identified in 1 case. One patient could not be thoroughly treated owing to a distal esophageal stricture and a small ulcer at the top. Hyperalimentation therapy or feeding through a jejunostomy tube inserted during operation was performed in all the surgically treated patients.

Table 1 Causes of perforation and means of treatment.

Cause	n	Nonsurgical therapy	Surgical drainage	Esophagectomy
Esophageal dilation	14	10	2	2
	3	1	1	1
Sclerotherapy	1	0	1	0
Gastroscopy	6	3	1	2
Blunt trauma	1	1	0	0
Barogenic	5	1	1	3
Spontaneous	1	1	0	0

RESULTS

The therapeutic efficiency was judged on the basis of the following clinical criteria. (1) Effective: the perforation healed without dysphagia. (2) Improved: the perforation closed with persistent dysphagia while swallowing solid food. (3) Ineffective: the patient died after therapy. Means and results of the therapy are shown in Table 2.

Table 2 Means and results of therapy.

Result	Nonsurgical therapy	Surgical drainage	Esophagectomy
Effective	9	2	5
Improved	6	3	0
Ineffective	2	1	3

Sixteen patients were cured, among whom 9 received nonsurgical therapy, 2 surgical drainage, and 5 esophagectomy. Nine patients had closure of the perforation that was accompanied by persistent dysphagia while swallowing solid food; among these patients, 6 received nonsurgical therapy and 3 received surgical drainage. Six patients died in total, with 2 from massive hemorrhage from the large collateral vessels near the gastroesophageal junction during operation, 1 from postoperative sepsis at 2 wk after the esophagectomy, 1 from suffocation resulting from enlargement of the perforation, 1 from infection, and 1 from arrythmia during surgical drainage. The cure rate was 51.6% (16/31) and the mortality rate was 19.4% (6/31).

DISCUSSION

Esophageal perforation is an important complication of esophageal endoscopy and dilation, which commonly occurs in the aged. It has been reported that esophageal dilation is the most frequent cause of esophageal perforation[3-5], and that the most frequent site of perforation is the thoracic esophagus. In our study, however, perforation caused by fiberoptic endoscopy was seen in only 1 case, in which a high cervical perforation of a Zenker′s diverticulum occurred during blind intubation for endoscopic retrograde cholangiopancreatography. The esophageal perforation rate due to diverticulum has been reported as 2% to 15%[6,7]. We performed endoscopy for 335 patients, and esophageal perforation only occurred in 1.8% (6/335). Ninety-seven patients were treated by esophageal dilation for benign stricture in our study, and the perforation rate was 14.4% (14/97). Barogenic perforation refers to esophageal disruption caused by an abrupt rise in esophageal pressure. In our study, acute vomiting was the most common cause of barogenic perforation in some healthy patients. One patient had non-barogenic spontaneous perforation. Pain of the chest, dysphagia, and shortness of breath were the most frequent symptoms. However, it is noteworthy that even in the cases of acute perforation after dilation, some patients did not have pain, and instead they complained of dysphagia on swallowing. It is important to recognize that perforation may occur within a short time after examination.

The temporary asymptomatic condition, perhaps, is due to edema at the site of a small perforation. Therefore, the patients should be presumptively treated only if a perforation is suspected. Foley et al[8] reported on 6 patients in whom water-soluble contrast media failure to display the perforation and subsequently received barium examination. Our study also demonstrated that perforation of the esophagus can be displayed with barium. Some experts have recommended that the initial contrast examination be performed with a water-soluble contrast agent, because it is thought to cause less irritation of the mediastinum. If the result is negative, barium meal examination should be followed[8]. Others recommended the use of thin barium as a selective agent[9]. There were two patients who each had an esophageal perforation that was confirmed by roentgenograms, but the results of the initial examination were negative. Thus, a repeated examination after several hours is important. Esophagoscopy can be used to identify the site of the perforation[3], but it is quite difficult to find the site of the perforation if a contrast agent is not applied before the endoscopic examination. In addition, insufflation of air may increase contamination of the mediastinum. Furthermore, some perforations may appear small, but they may again become open perforations requiring surgical treatment if medical treatment is given only.

Generally, the symptoms of perforation are very typical. The low mortality in our series of patients was in part related to early investigation of the patients with dysphagia, chest pain, or shortness of breath after dilation. The endoscopist often stands for conservative treatment for esophageal perforation caused by endoscopy, which may bungle the chance of surgical treatment. Delayed treatment of more than 24 h may cause more dangerous complications and led to a higher mortality rate. In our study, surgical treatment was immediately performed in the patients who did not meet the criteria for nonsurgical management (within 24 h in the majority of the cases), and no patient initially treated with medical treatment required subsequent surgery.

The rates of mortality from esophageal perforation ranged from 7% to 47%[4,10]. In a collective review of 450 cases, Jones and Ginsberg[9] reported that the mortality of patients with iatrogenic or instrumental perforation was 19%, whereas those with spontaneous perforations had a mortality of 39%. Two other studies[1,12], however, showed no difference between spontaneous and instrumental perforations. In our study, the higher mortality in the surgical group was attributable to serious contamination of the mediastinum in those patients with large perforations, delayed diagnosis and life-threatening complication, which should have manifested symptoms in the course of medical treatment.

The primary closure of esophageal perforation has a poor outcome, particularly in patients with a distal esophageal stricture, and some surgeons recommended more definitive surgery for the stricture in early or late stage[3]. The second operation on patients with dysphagia must be determined according to the condition of each patient. Among the two patients undergoing primary closure in our hospital in 1987, perforation recurred in 1 patient and another had aggravated dysphagia (unable to take liquid); both patients underwent another esophagectomy. In the follow-up of our patients, we found that some patients refused surgery despite their persistent dysphagia, and a small number of the patients preferred surgical treatment to dilation. As a matter of fact, esophageal dilation is liable to lead to a perforation in the aged because of degenerative changes in the tissues of the esophagus. Early diagnosis is not difficult because of its typical symptoms, but esophageal dilation should be improved.