CURRENT REVIEW

Esophageal Perforation: A Continuing Challenge William G. Jones 11, MD, and Robert J. Ginsberg, MD Division of Thoracic Surgery, Memorial Sloan-KetteringCancer Center, and Division of Cardiothoracic Surgery, Cornell University Medical College, New York, New York

Perforation of the esophagus remains a diagnostic and therapeutic challenge. Currently, the most common cause of perforation is instrumentation of the esophagus, and the incidence of esophageal perforations has increased as the use of endoscopic procedures has become more frequent. Diagnosis depends on a high degree of suspicion and recognition of clinical features, and is confirmed by contrast esophagography or endoscopy. Outcome after esophageal perforation is dependent on the cause and location of the injury, the presence of

underlying esophageal disease, and the interval between injury and initiation of treatment. Reinforced primary repair of the perforation is the most frequently employed and preferable approach to the surgical management of esophageal perforations. Nonoperative management consisting of antibiotics and parenteral nutrition is particularly successful for limited esophageal injuries meeting proper selection criteria.

D

and by Olsen and Clagett [6] in 1947. In that same year, Brewer and Burford [7] also reported the successful surgical repair of a traumatic esophageal perforation. Five years later, Satinsky and Kron [8] reported the first successful esophagectomy after a perforation. Over the subsequent 40 years, the incidence of esophageal perforation has increased [9], iatrogenic causes of esophageal perforations have become more prevalent than spontaneous ruptures, and different management techniques have evolved.

espite the many advancements made in thoracic surgery, the management of patients with esophageal perforations remains problematic and controversial. Although results have improved since the advent of broad-spectrum antibiotics and better nutritional support, esophageal perforation still represents a continuing challenge to the clinician. The diagnosis is often difficult to make, as clinical findings may be subtle and mimic other disorders, yet delay in diagnosis and initiation of treatment results in potentially high morbidity and mortality. Further, the low frequency of occurrence of esophageal perforation leads to a lack of experience for most individual surgeons. In addition, no single method of therapy is universally agreed on as ideal, and the frequent presence of underlying esophageal disease complicates management.

Historical Notes The clinical features of esophageal perforation have long been recognized and have been extensively described in the literature. Although mention is made of an esophagocutaneous fistula secondary to perforation in the Smith papyrus dating from 2,500 BC [l], the first report of spontaneous esophageal perforation was made by Boerhaave [2] in 1724. In 70 pages of clinical and pathological observations, Boerhaave described the distal esophageal rupture that occurred in the High Admiral Baron de Wessenaer in association with forceful vomiting after overindulgence in food and drink. Although Frink [3] had treated a patient with esophageal rupture by surgical drainage in 1941, it was not until 1944 that the first surgical repair after spontaneous perforation was attempted by Collins and associates [4], laying the foundation for the first reported successful repairs by Barret [5] Address reprint requests to Dr Ginsberg, Division of Thoracic Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, New York, NY 10021.

0 1992 by The Society of Thoracic Surgeons

(Ann Thorac Surg 1992;53:53443)

Etiology and Location The causes of esophageal perforation are as follows:

I. Instrumentatiodiatrogenic A. Intraluminal 1. Esophagoscopy 2. Bougienage 3. Pneumatic dilation 4. Sclerosis of esophageal varices 5. Placement of intraesophageal tubes (nasogastric, Sengstaken-Blakemore, prostheses) 6. Endotracheal tube placement B. Extraluminal 1. Mediastinoscopy 2. Intraoperative injury a. Thyroid resection b. Leiomyoma enucleation c. Proximal gastric vagotomy d. Pneumonectomy 3. Radiation therapy 11. Traumatic

A. Blunt B. Penetrating C. Ingestion of caustic substance 0003-4975/92/$5.00

REVIEW JONES AND GINSBERG ESOPHAGEAL PERFORATION

Ann Thorac Surg 1992;5353443

'-1 80%

-

60%

35%

Fig 1. (A) Cause of 51 1 cases of esophageal perforation in recent series, and ( B ) types of instrumental perforations among those cases. (Data from references 49-51, 55, 72, 112, and 113.)

111. Spontaneous (barogenic) A. Postemetic B. Other IV. Ingestion of foreign body V. Tumor VI . Surrounding infections Their relative frequency from reported series is depicted in Figure 1, and the relationship of cause to location is shown in Figure 2. Instrumental esophageal perforation occurs in about 1/10,000 upper gastrointestinal endoscopies. In a 1974 American Society of Gastrointestinal Endoscopy survey [lo] of 211,410 procedures, perforation occurred in 0.03%

535

of all simple upper endoscopies and in 0.25% of procedures involving bougienage. With an incidence ranging from 1%to 10% [ l l , 121, pneumatic dilation for achalasia is associated with a fourfold greater risk of perforation than myotomy [13], and this may be related to patient malnutrition [141 or high-compliance balloons [151. Endoscopic sclerosis of esophageal varices leads to perforation in 1% to 4% of cases [16-191 and the result may be in part dependent on the type and volume of sclerosing agent employed [18, 191, although this concept is disputed by Korula and colleagues [20]. Esophageal perforation can also occur during placement of nasogastric tubes [21, 221, Sengstaken-Blakemore tubes [23, 241, endoesophageal prostheses [25-271, and endotracheal tubes [28-301. The area at greatest risk for instrumental injury is the cricopharyngeal region of the cervical esophagus. At Lannier's triangle, formed by the constrictor pharyngeus and the cricopharyngeus muscles at the level of the C-5 and C-6 vertebrae, the posterior esophageal mucosa is covered only by fascia. The risk of perforation in this area is further increased by hyperextension of the neck and the presence of kyphosis or cervical vertebral spurs [31]. The next most common areas of instrumental esophageal perforation are the relatively narrow but fixed portions at the distal esophagus just proximal to the hiatus and near the aortic arch and left main bronchus. Other areas at risk include portions of the esophagus just proximal to obstructing processes, portions with extensive carcinomas, and portions at biopsy sites. Intraoperative injury to the esophagus can occur in areas where an operative procedure is in close proximity to the esophagus. Intraoperative esophageal injury has been reported after anterior cervical spine operations [32, 331, mediastinoscopy [34], thyroid resection [l],leiomyoma enucleation [11, proximal gastric vagotomy [35-381, hiatal hernia repair [35,39,40], and pneumonectomy [41]. Perforations from either blunt or penetrating trauma occur primarily in the neck, whereas perforations from ingested foreign bodies or caustic substances are more likely to be intrathoracic. Spontaneous esophageal perforations are barogenic in origin, resulting from a sudden rapid rise in intraesophageal pressure such as can occur with forceful vomiting, childbirth, weight lifting, or blunt abdominal trauma. Barogenic perforations usually occur on the left wall of the distal supradiaphragmatic esophagus, extending into the pleural cavity in 80% of cases [I, 42, 431. Other less common causes such as perforations of esophageal carcinomas [4447] and perforations from surrounding infections, especially in immunocompromised patients [48], can occur anywhere in the intrathoracic esophagus. The association of underlying esophageal disease with esophageal perforation is depicted in Figure 3. Among 147 patients in seven large series of esophageal perforation from a variety of causes [23,4%54], no underlying esophageal pathology was present in 40%.Benign strictures and anatomic lesions such as diverticula were most commonly associated with perforations, usually instrumental in origin, whereas tumors or achalasia occurred less frequently.

536

REVIEW JONES AND GINSBERG ESOPHAGEAL PERFORATION

Fig 2 . Relationship of cause to location of esophageal perforations in recent series. (Data from references 49, 51, 55, 112, and 113.)

Ann Thorac Surg 1992;5353443

e0 5n

-

loo%]

Operative injury Spontaneous Trauma instrument

Clinical Features Diagnosis of esophageal perforation can be difficult, as the presentation is often nonspecific and is easily confused with other disorders such as peptic ulcer disease, pancreatitis, myocardial infarction, dissecting aortic aneurysm, pneumonia, or spontaneous pneumothorax [55, 561. The signs and symptoms of esophageal perforation depend on the location and cause of the perforation and the interval since the perforation occurred. Cervical perforations most often occur posteriorly where the esophageal wall is thin. Dissection through the retroesophageal space allows spread of contamination to the mediastinum over time, but attachments of the esophagus to the prevertebral fascia may limit lateral spread. Thus, early signs of cervical esophageal perforation include neck stiffness and a dull neck ache, regurgitation of bloody material, and the finding of cervical subcutaneous lW%

-

80%

-

60%

I-

40%

Fig 3 . Underlying esophageal disease in patients with esophageal perforation. (Data from references 23 and 49 through 54.)

emphysema. Inflammatory changes in the neck may not develop for several hours, with signs of systemic sepsis often not occurring for up to 24 hours. In contrast, perforations of the thoracic esophagus result directly in mediastinal contamination, leading to a more rapid development of pneumomediastinum and mediastinitis than after cervical perforations. The thin mediastinal pleura is usually ruptured by the inflammatory process, producing contamination of the pleural space and a pleural effusion. Gastric contents and fluids are then drawn into the pleural space by the negative intrathoracic pressure [41],resulting in further inflammation and fluid sequestration, hypovolemia, and the early appearance of tachycardia and systemic sepsis. Chest pain and subcutaneous emphysema are usually present, and dyspnea is often prominent even in the absence of pneumothorax. Intraabdominal esophageal perforations occur into the free peritoneal cavity and result in peritonitis. A dull retrosternal ache in association with epigastric pain radiating to the shoulders is characteristic because of the relationship of the intraabdominal esophagus to the diaphragm. Systemic signs such as tachycardia, tachypnea, and fever develop early, with progression to sepsis and shock occurring within hours.

Diagnosis Prompt diagnosis of esophageal perforations is essential, as the incidence of complications and the rate of mortality are dependent on the interval between the perforation and the initiation of treatment. The presence of neck, chest, or abdominal pain after upper gastrointestinal endoscopy should always raise the suspicion of an esophageal perforation and necessitates further study to eliminate the possibility that perforation has occurred. Plain film findings vary according to the location and cause of the perforation, as well as the interval between injury and radiographic examination [57]. Findings of cervical or mediastinal emphysema, pneumothorax or pneumopericardium, pleural effusion, or subdiaphrag-

REVIEW JONES AND GINSBERG ESOPHAGEAL PERFORATION

Ann Thorac Surg 199253353443

matic air are all suggestive of a perforated esophagus and warrant further diagnostic study with contrast esophagograms. The presence of food particles or a pH of less than 6.0 in the pleural fluid aspirate is virtually diagnostic [58]. Noncontrast radiographic studies will be normal in 12% [57] to 33% [59]of cases of esophageal perforation, leading some authors [60] to recommend contrast studies after all endoscopic dilations. Although contrast esophagograms are the standard diagnostic procedures in cases of suspected esophageal perforation, it should be noted that the false-negative rate of these examinations can also exceed 10% [45]. Thus, "negative" studies may not completely eliminate the possibility of a perforation. When positive, contrast studies have the advantage of demonstrating the level of the perforation and the presence of extension into the pleural cavity. The ideal contrast agent for esophagography for suspected perforation was previously the subject of some debate [61-641. Water-soluble contrast medium is rapidly absorbed and is purported to be less irritating than barium should substantial extravasation occur [65]. Our preference is thinned barium, which may improve localization of subtle perforations [66], and it should be employed whenever a major risk of aspiration is present or whenever a concomitant respiratory tract injury or fistula is suspected. Foley and co-workers [65] recommend an initial bolus of water-soluble contrast medium and then a repeat of the study with a small amount of barium if no leak is seen. In the patient in whom esophageal perforation is highly suspected clinically but in whom contrast esophagograms are negative, flexible esophagoscopy and computed tomography can be useful diagnostic adjuvants. Flexible esophagoscopy may currently be underutilized in the diagnosis of esophageal perforation. This method provides highly accurate localization as well as direct visualization of perforations [67]. Furthermore, results of esophagoscopy may directly influence treatment decisions, as small perforations or perforations limited to the mucosa can be expected to heal with nonoperative management in the absence of distal obstruction. In select cases, esophagoscopy can facilitate placement of a tube through the perforation for irrigation and drainage [68]. Computed tomographic findings suggestive of esophageal perforation include air in the soft tissues of the mediastinum surrounding the esophagus, abscess cavities adjacent to the esophagus in either the pleural space or the mediastinum, or the demonstration of an actual communication between the air-filled esophagus and an adjacent mediastinal or paramediastinal air-fluid collection [58]. Computed tomographic scans are also particularly useful in follow-up after initiation of therapy and in the evaluation of patients who fail to improve despite either operative or nonoperative management [69]. Localization of loculated pleural fluid collections can also allow computed tomography-directed drainage catheter placement in the postoperative setting or during nonoperative ther~ P [701. Y

537

Treatment of Esophageal Perforations Once the diagnosis of esophageal perforation is established, multiple factors must be considered in selecting the appropriate therapy. The cause of the perforation, its location, the presence of underlying esophageal disease, and the interval between perforation and diagnosis are critical factors in the determination of treatment. In addition, the condition of the esophagus and the extent of soilage or injury to adjacent organs and tissues as well as the age and general condition of the patient must also be considered. Treatment options for esophageal perforations are as follows: 1. Operative management a. Primary closure b. Reinforced primary closure c. Resection d. Drainage alone e. T-tube drainage f. Exclusion and diversion g. Intraluminal stent 2. Nonoperative management The goal of any treatment modality must be to prevent further soilage from the perforation, to eliminate the infection produced by soilage, to restore the integrity and continuity of the gastrointestinal tract, and to restore and maintain adequate nutrition [26, 45, 71-74]. Consequently, the principles of operative management include debridement of all infected and necrotic tissue, secure closure of the perforation with appropriate measures to prevent leakage, correction or elimination of distal obstruction, and drainage of contaminated and infected areas. As recovery from esophageal perforation can be prolonged, provision of a route for enteral nutrition such as a jejunostomy should also be considered during any surgical procedure.

Surgical Approaches The surgical approach is dependent on the location of the perforation. Cervical esophageal tears are best exposed through an incision parallel to the left sternocleidomastoid muscle and anterior to the carotid artery and internal jugular vein. Perforations in the upper two-thirds of the thoracic esophagus can be reached through a right posterolateral thoracotomy in the fourth or fifth intercostal space, whereas those in the lower third are best approached through a posterolateral thoracotomy in the left sixth or seventh interspace. Abdominal esophageal injuries usually require an upper midline laparotomy. Once the esophagus is exposed, localization of subtle perforations may require instillation of methylene blue into the esophageal lumen or the insufflation of air into the esophagus immersed in saline solution.

Primary Closure Primary closure of esophageal perforations was first described nearly 50 years ago [4-61. Primary closure, with or

538

REVIEW JONES AND GINSBERG ESOPHAGEAL PERFORATION

without tissue reinforcement, is the most frequently used and preferable approach to the surgical management of esophageal perforations. Successful repair depends on debridement of nonviable tissue and closure of the mucosa and muscle in separate layers. Closure of the mucosa with mechanical stapling devices has also been reported [75, 761. In most instances, the mucosal layer remains healthy. However, a myotomy may be necessary to visualize the full extent of the mucosal injury and to facilitate its repair [75, 771. Inadequate exposure and repair of the mucosal defect can predispose to leakage from the closure

PI.

Late presentation or delayed diagnosis of the perforation results in further tissue necrosis and edema, and can prevent closure or even identification of a substantial muscular layer, in which case the mucosa alone is closed. The contaminated thoracic cavity should be thoroughly irrigated and debrided, with decortication of the lung if necessary. Wide external drainage consisting of Penrose drains in the neck for cervical injuries and large-bore chest tubes for thoracic esophageal perforations should be established along with nasogastric drainage. Contrast studies are usually obtained 7 to 10 days postoperatively, at which time oral intake can be resumed if no leak is demonstrated and ileus has resolved. Primary closure is best applied to perforations in an otherwise normal esophagus. In the case of prestenotic esophageal perforation, closure alone without resolution of the obstructing process can be doomed to failure [78, 791. Moghissi and Pender [67] reported a 100% mortality rate among 9 patients with prestenotic perforations treated by either drainage alone or primary closure. For the 14 other patients in the series in whom the stenosis as well as the perforation was addressed, there was a reduction in the mortality rate to 29%.

Reinforced Closure The potential for breakdown or leakage from primary closure of esophageal perforations, particularly if the interval between injury and treatment exceeds more than a few hours, has led many surgeons to advocate reinforcement of the closure. In the chest, a variety of tissues have been employed to buttress the esophageal closure. A pleural flap, as suggested by Grillo and Wilkins [80], is particularly useful if local reaction has resulted in a thickened pleura. Other choices include omentum, a pedicled intercostal muscle flap [81, 821 or pericardial fat pad flap [83, 841, elevation of a flap of diaphragm [85,86], or coverage by infolded gastric wall as a Thal patch [87, 881 or as part of an antireflux procedure if enlargement of the hiatus is necessary or underlying reflux disease is present [56]. Regardless of the type of material used, secure closure necessitates suturing the reinforcing tissue closely, as for an anastomosis, rather than just simple “tacking” [26].

Esophageal Resection Excellent results with esophageal resection, particularly for perforations in association with malignant obstruction or massive esophageal necrosis, have been reported [46].

Ann Thorac Surg 1992:53:53443

A transhiatal or thoracic approach can be employed, with reconstruction performed immediately or as a secondstage procedure depending on the degree of contamination and the condition of the patient. Matthews and associates [89] advocate the use of near-total esophagectomy if resection is required. They cite the ease of performing the anastomosis in the neck, the decreased necessity to use inflamed tissues in the anastomosis, and better longitudinal lymphatic clearance in support of their view.

Concomitant Treatment of Underlying Esophageal Disease Perforation in association with functional obstruction, as in achalasia, is best treated by repair with a myotomy performed on the side of the esophagus opposite the perforation. An antireflux procedure is generally not required if the myotomy does not extend onto the gastric cardia [90]. However, a Belsey-type repair may offer the advantage of buttressing the closure of distal perforations [91]. An antireflux procedure should also be performed when perforation is associated with hiatal hernia or major reflux esophagitis.

Thoracotomy and Drainage Drainage alone, without repair of the perforation, is sometimes deemed necessary in the face of widespread inflammation. This technique is best applied to cervical perforations, where better drainage of the surrounding tissues can be obtained than in the chest, as it does not prevent further soilage. Closure of the perforation around a 6- to 10-mm T tube, as reported by Abbott and coworkers [92], creates a controlled esophagocutaneous fistula “venting” the esophagus. This technique may be useful when the security of an attempted repair is in doubt or adequate tissue to reinforce the closure is not available. Intraluminal tubes or stents have also been used successfully to treat perforations through carcinomas or in association with tracheoesophageal fistulas [93-961.

Exclusion and Diversion Urschel and others [74, 97-99] have described exclusion/ diversion techniques in the treatment of complicated esophageal perforations. Exclusion and diversion involves suture of the perforation, if possible, with wide drainage of contaminated tissues; creation of a loop or end-cervical esophagostomy; division or suture of the gastroesophageal junction; and creation of a gastrostomy. Advocates of these techniques claim that they are an application of principles used successfully in the treatment of perforations of other areas of the gastrointestinal tract such as perforated colonic diverticulitis [97]. Exclusion and diversion, however, commits the patient to a second major operative procedure and can create a stagnant pool in the distal esophagus with bacterial overgrowth. Gouge and associates [50] wrote that the “rationale for [the] procedure is debatable and depends on whether the benefits of preventing reflux from the distal esophagus outweighs the disadvantage of creating an obstruction in the distal esophagus.” Thus, this technique

Ann Thorac Surg 1992;5353443

should be reserved for those patients in whom other approaches are not feasible.

Nonoperative Management In the era before antibiotics, nonoperative treatment of esophageal rupture was almost universally fatal [loo]. More recently, Michel and co-authors [26] considered the findings of pneumothorax, pneumoperitoneum, pneumomediastinum, systemic sepsis, shock, or respiratory failure to be indications for operative management. These concepts are being challenged, however, as more patients with esophageal perforation are being treated without surgical intervention. The justification for this deviation from time-honored surgical therapy is fourfold: (1) most perforations are now the result of instrumentation and therefore are usually more limited, associated with less contamination, and diagnosed more promptly than spontaneous ruptures; (2) newer, more effective antibiotics have been introduced to control the infections that formerly resulted from esophageal perforation; (3) computed tomographic imaging provides the ability to more adequately drain the thoracic cavity through a percutaneous approach; and (4) there are now safe and effective forms of parenteral nutrition and alternative means of enteral nutrition. Opponents of nonoperative management recommend primary closure to decrease recovery time and minimize time of fasting and total parenteral nutrition [1011. Nonoperative therapy is best applied in the following cases: (1) instrumental perforation, especially in the cervical esophagus; (2) small perforations after bougienage of peptic strictures or achalasia or after sclerosis of esophageal varices, as the presence of periesophageal fibrosis can limit contamination of the mediastinum; and (3) esophageal perforation diagnosed several days after injury with minimal symptoms [102-1081. Nonoperative therapy is likely to fail in the presence of esophageal obstruction [109]. Based on their 30-year experience, Berry and Ochsner [1101 recommended an individualized approach to the decision of operative versus nonoperative therapy. Cameron and associates [lo21 and others (70, 103-1081 have established criteria for nonoperative management of esophageal perforations. These criteria are as follows: 1. Recent perforation or perforation well walled off if seen late 2. Esophageal perforation well contained within the mediastinum or between the mediastinum and visceral lung pleura, with no free contrast agent escaping into adjacent body cavities 3. Cavity well drained back into the esophagus with minimal pleural soilage 4. No food intake between injury and diagnosis 5. Perforation not in area of tumor or in the abdomen or proximal to an obstructing process 6. Minimal symptoms 7. Minimal signs of clinical sepsis and minimal physiologic derangement Patients are given nothing by mouth for usually more than 10 days, necessitating total parenteral nutrition.

REVIEW JONES AND GINSBERG ESOPHAGEAL PERFORATION

539

Broad-spectrum antibiotic therapy is administered for a minimum of 7 to 14 days. Pleural collections are drained through chest tubes or computed tomography-directed drainage catheters [70]. Nasogastric drainage is widely considered integral to nonoperative therapy, although some authors [lo21 report using nasogastric tubes only selectively. Nasogastric tubes placed in the region of the perforation can also be used for irrigation with either saline or antibiotic solutions as an adjunct to nonoperative treatment [68]. Failure to show signs of improvement within 24 hours of nonoperative therapy should prompt a consideration for operative intervention [25]. Computed tomographic scanning may be helpful to demonstrate undrained collections in such cases [l, 691.

Outcome Recent series [49, 51, 721 have identified cause, location, delayed treatment, underlying esophageal disease, and type of treatment as significant risk factors for death after esophageal perforation. Examination of reported series of esophageal perforations published from 1980 through 1990 [23, 4547, 49-53, 67, 68, 71, 111-1131, the era of modern antibiotic therapy and parenteral nutrition, allows an estimate of the current risk associated with each of these factors. Among 450 reported patients [45, 47, 49, 51, 55, 67, 111-1131, iatrogenic and instrumental injuries were associated with a mortality rate of 19%, whereas spontaneous ruptures resulted in a mortality rate of 39%.Two individual series [114, 1151, however, comparing outcome after treatment for spontaneous and iatrogenic perforations failed to demonstrate any significant difference in associated mortality between these two causes. Traumatic esophageal perforations in the collected series [45, 47, 49, 51, 55, 67, 111-1131 had the lowest mortality rate, 9%, perhaps because of a decreased interval between perforation and treatment in the setting of multiple injury [113]. In 439 patients [7, 49, 51, 55, 67, 1131, cervical tears resulted in an overall mortality rate of 6%, whereas thoracic and abdominal perforations resulted in death in 34% and 29% of patients, respectively. The diminished risk associated with perforation of the esophagus in the neck is attributed to anatomic tissue planes in this area, which prevent the rapid spread of contamination or infection [l], and to the absence of reflux of gastric contents through cervical perforations [116]. Prompt recognition and initiation of treatment of esophageal perforations has long been recognized as essential to a favorable outcome [117]. Delay in treatment of greater than 24 hours after perforation has occurred has been associated with higher rates of both complications [43] and mortality [43, 49, 51, 55, 67, 721. Bladergroen and associates [49] reported that survival decreased from 79% to 67% and Attar and colleagues (511, from 87% to 55% if treatment was instituted more than 24 hours after the onset of symptoms. Basing a treatment plan on the interval between perforation and presentation has also been associated with reduction in mortality [72]. However, the influence of treatment delay may have dimin-

540

REVIEW JONES AND GINSBERG ESOPHAGEAL PERFORATION

Ann Thorac Surg 1992:53:53443

Table 1 . Outcome After Treatment of Esophageal Perforation in Series Published Between 1980 and 1990"

Reference

Year

Skinner et a1 [47] Goldstein and Thompson [112]

1980 1982 1982 1983 1984 1984 1986 1986 1987 1988 1989 1989 1990

Sarr et a1 [45]

Larsen et a1 [23] Ajalat and Mulder [ l l l ] Borgeskov et a1 (531 Radmark et a1 [52] Brewer et a1 [71] Nesbitt and Sawyers [55] Moghissi and Fender [67] Flynn et a1 [113] Gouge et a1 [50] Attar et a1 [51] Total

Mortality (%)

Primary Repair 0115 4/23 0115' 10147 0112 9/22 Vlrl 9/53 3/20 7/13' 1/44 0114 5/30 491325 15

Exclusion Drainage 48

... 0113 48 115 317

... 016 314 45 2/9

011 7/17 28/83 34

and

Overall

Resection

Diversion

Nonoperative

Mortalityb

319

2/11 619 013

...

9/43 (21) 14/44 (32) 4/47 (9) 14/57 (25) 5/33 (15) 19/39 (49) 7/38 (18) 9/78 (12) 8/34 (24) 13/29 (45) 5/67 (7) 3/18 (17) 18/61 (30) 1281589 22

... 318

... 011 2/2 012 012

... a 11 114 111 2/9 14/49 29

... 112

... 112 012 118

... 112 2/2 4/5 18/46 39

4/12 118 012 3/13 518 5/17 0115 112

... 018

... ...

19/85 22

Data are presented as number of deathdnumber of patients undergoing procedure. Numbers in parentheses are percentages. This includes 2 patients who also underwent an antireflux procedure. This includes 2 patients who also underwent myotomy. This includes 5 patients who also underwent fundoplication. a

ished slightly. Nesbitt and Sawyers [55] noted that if treatment was initiated within 24 hours, the mortality rate decreased from 56% to 13%during the years 1935 through 1974 but only decreased from 26% to 11.4% in the years 1975 through 1984. Decreases in mortality in patients with delayed diagnosis and treatment after esophageal perforation may be attributable to improvements in critical care, antibiotic therapy, and parenteral nutrition. Patients with cancer of the esophagus regardless of the site of perforation appear to have increased mortality compared with patients with perforations associated with benign disease processes or no underlying esophageal disease [26, 49, 72, 1141. Further, Saaybe and co-workers [118] studied the long-term results in patients with esophageal perforations treated by repair without further surgical treatment of any underlying benign esophageal disease. They found that all patients without underlying esophageal disease remained asymptomatic up to 22 years after repair of instrumental perforations, whereas 35% of those with underlying disease had symptoms requiring reinstrumentation or reoperation. Table 1 depicts the outcome after treatment of esophageal perforation in 13 series published between 1980 and 1990 [23, 45, 47, 50-53, 67, 71, 111-1131 in which the mortality for individual procedures is reported. Among 325 patients treated with various types of primary repairs, the mortality rate ranged from 0% to 54% and averaged 15%. Gouge and associates [50] reported a 25% mortality in 158 personal and reported patients treated with primary repair versus a 6% mortality in 99 patients treated with reinforced repair. The rate of fistula formation or serious leak also decreased from 39% to 13% when the esophageal closure was reinforced. Other forms of operative management are associated with higher mortality

rates in these series: exclusion and diversion, 39%; drainage only, 34%; and resection, 29%, although these increased mortality rates may also be a reflection of more complicated disease. Nonoperative therapy resulted in a 22% mortality rate in these series (see Table 1).Previous reviews of series of esophageal perforations have reported a 25% mortality rate among 900 patients [112] in 1982 and a 38% rate among 428 patients in 1983 [23]. However, it is difficult among these cases to distinguish between election of nonoperative management as '%best therapy" and nonoperative treatment because of inability to tolerate or refusal to undergo operative therapy, as multiple anecdotal reports and small series of instrumental and pediatric perforations have survival rates approaching 100% [68, 70, 103-1081. Erwall and associates [119] reported similar mortality rates between patients treated nonoperatively (12%)and those treated operatively (16%)in a series of 48 patients in whom the presence of shock or systemic symptoms, large perforations, retained foreign bodies, or esophageal malignancy were considered operative indications. Thus, patient selection according to the strict criteria listed in the Outcome section would appear to be necessary for nonoperative treatment to be successful.

Conclusion Esophageal perforation is a potentially lethal condition resulting most frequently from instrumentation. Early diagnosis and intervention are necessary to prevent morbidity and mortality. Esophageal perforation is best treated with definitive repair including tissue reinforcement of the closure and elimination of distal obstruction.

Ann Thorac Surg 1992;53:534-43

Nonoperative therapy consisting of antibiotics and parenteral nutrition may also be appropriate in select patients.

References 1. Witte J, Pratschke E. Esophageal perforations. In: Baue AE, Geha AS, Hammond GL, Laks H, Naunheim KS, eds. Glenn's thoracic and cardiovascular surgery. 5th ed. Norwalk, CT: Appleton & Lange, 1991:669-78. 2. Boerhaave H. Atrocis, nec descripti prius, morbi historia secundem artis leges conscripta, lugduni batavorum, bontes teniana. Medici 1724:60. 3. Frink NW. Spontaneous rupture of the esophagus: report of a case with recovery. J Thorac Surg 1941;16291-2. 4. Collins JL, Humphreys DR, Bond WH. Spontaneous rupture of the esophagus. Lancet 1944;2:17944. 5. Barrett NR. Report of a case of spontaneous rupture of the esophagus successfully treated by operation. Br J Surg 1947;35:216-7. 6. Olson AM, Clagett OT. Spontaneous rupture of the esophagus. Report of a case with immediate diagnosis and successful surgical repair. Postgrad Med 1947;2:417-9. 7. Brewer LA, Burford TH. Special types of thoracic wounds. In: Medical Department: US Army Surgery in World War 11, Thoracic Surgery. 38th ed. Washington, DC: Government Printing Office, 1965:269. 8. Satinsky VP, Kron SD. One-stage esophagectomy in the presence of mediastinitis. Arch Surg 1952;64:1247. 9. Sawyers JL, Lane CE, Foster JH, Daniel RA. Esophageal perforation: an increasing challenge. Ann Thorac Surg 1975; 19:233-8. 10. Silivis SE, Nebel 0, Rogers G, Sugava C, Mandelstam P. Endoscopic complications: result of the 1974 American Society of Gastrointestinal Endoscopy survey. JAMA 1976;235: 928-30. 11. Okike N, Payne WS, Neufeld DM, Bernatz PE, Pairolero PC, Sanderson DR. Esophagomyotomy versus forceful dilation for achalasia of the esophagus: results in 899 patients. Ann Thorac Surg 1979;28:119-25. 12. Miller RE, Bossart PW, Tiszenkel HI. Surgical management of complications of upper gastrointestinal endoscopy and esophageal dilation including laser therapy. Am Surg 1987; 53:667-71. 13. Sauer L, Pelligrini CA, Way LW. The treatment of achalasia: a current perspective. Arch Surg 1989;124929-32. 14. Fennerty MB. Esophageal perforation during pneumatic dilatation for achalasia: a possible association with malnutrition. Dysphagia 1990;5:227-8. 15. Rabinovici R, Katz E, Goldin E, Kluger Y, Ayalon A. The danger of high compliance balloons for esophageal dilatation in achalasia. Endoscopy 1990;22:634. 16. Bacon BR, Camara DS, Duffy MC. Severe ulceration and delayed perforation of the esophagus after endoscopic variceal sclerotherapy. Gastrointest Endosc 1987;33:3114. 17. Schumann BM, Beckman JW, Tedesco FJ. Complications of endoscopic injection sclerotherapy: a review. Am J Gastroenterol 1987;82:82&3-30. 18. Shemesh E, Bat L. Esophageal perforation after fiberoptic endoscopic sclerotherapy for esophageal varices. Arch Surg 1986;121:243-7. 19. Perino LE, Gholson CF, Goff JS. Esophageal perforation after fiberoptic variceal sclerotherapy. J Clin Gastroenterol 1987;9:286-9. 20. Korula J, Pandya K, Yamada S. Perforation of esophagus after endoscopic variceal sclerotherapy. Incidence and clues to pathogenesis. Dig Dis Sci 1989;34:324-9.

REVIEW JONES AND GINSBERG ESOPHAGEAL PERFORATION

541

21. James RH. An unusual complication of passing a narrow bore nasogastric tube. Anaesthesia 1978;33:716-8. 22. Jackson RH, Payne K, Bacon BR. Esophageal perforation due to nasogastric intubation. Am J Gastroenterol 1990;85: 439-42. 23. Larsen K, Skov Jensen B, Axelsen F. Perforation and rupture of the esophagus. Scand J Thorac Cardiovasc Surg 1983;17311-6. 24. Rubin WC, Winsett MZ, Diner WC. Intrathoracic gastric balloon: radiographic recognition of esophageal perforations. Gastrointest Radio1 1982;7311-3. 25. Jansen CW. Perforation of the intrathoracic esophagus. Scand J Thorac Cardiovasc Surg 1976;10189-92. 26. Michel L, Grillo HC, Malt RA. Operative and nonoperative management of esophageal perforations. Ann Surg 1981; 1945743. 27. Hankins JR, Cole FN, Attar S, SatterfieldJR, McLaughlinJS. Palliation of esophageal carcinoma with intraluminal tubes: experience with 30 patients. Ann Thorac Surg 1979;28: 224-9. 28. Johnson KG, Hood DD. Esophageal perforation associated with endotracheal intubation. Anesthesiology 1986;64: 281-3. 29. Blair GK, Filler RM, Theodorescu D. Neonatal pharyngoesophageal perforation mimicking esophageal atresia: clues to diagnosis. J Pediatr Surg 1987;22:770-4. 30. Dubost C, Kaswin D, Durante A, Jehanno C, Kaswin R. Esophageal perforation during attempted endotracheal intubation. J Thorac Cardiovasc Surg 1979;78:4&51. 31. Seybold WD, Johnson MA, Learly WV. Perforation of the esophagus. Surg Clin North Am 1950;30:115!5-64. 32. Kelly MF, Spiegel J, Rizzo KA, Zwillenberg D. Delayed pharyngoesophageal perforation: a complication of anterior spine surgery. Ann Otol Rhino1 Laryngol 1991;100201-5. 33. Newhouse KE, Lindsey RW, 'Clark CR, Lieponis J, Murphy MJ. Esophageal perforation following anterior cervical spine surgery. Spine 1989;14:1051-3. 34. Demevik L, Larsson S, Pettersson G. Esophageal perforation during mediastinoscopy: the successful management of two complicated cases. J Thorac Cardiovasc Surg 1985;33: 179-80. 35. McBumey RP. Perforation of the esophagus: a complication of vagotomy or hiatal hernia repair. Ann Surg 1969;169: 851-3. 36. Postlethwait RW, Kim SK, Dillon ML. Esophageal complications of vagotomy. Surg Gynecol Obstet 1969;128:481-4. 37. Price JJ, Powis SJ, Morrissey DM. Oesophageal perforation during abdominal truncal vagotomy for duodenal ulcer. Br J Surg 1972;59:936-7. 38. Hauser JB, Lucas RJ. Esophageal perforation during vagotomy. Arch Surg 1970;101:466-7. 39. Skinner DB, Belsey RH. Surgical management of esophageal reflux and hiatal hernia: long-term results with 1030 patients. J Thorac Cardiovasc Surg 1967;53:3?-40. 40. Fletcher PR. Acute hiatal hernia with oesophageal perforation following Heller's operation. Br J Surg 1978;65:486-8. 41. Barrett NR. Discussion on unusual aspects of esophageal disease: perforation of the esophagus and of the pharynx. Proc R SOCMed 1956;49:529-32. 42. OConnell ND. Spontaneous rupture of the esophagus. AJR 1967;99:186-90. 43. Pate JW, Walker WA, Cole FH Jr, Owen EW, Johnson WH. Spontaneous rupture of the esophagus: a 30-year experience. Ann Thorac Surg 1989;47689-92. 44. Yeo CJ, Lillemoe KD, Klein AS, Zinner MJ. Treatment of

542

45. 46. 47.

48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59.

60.

61. 62. 63.

64. 65.

66. 67. 68.

REVIEW JONES AND GINSBERG ESOPHAGEAL PERFORATION

instrumental perforation of esophageal malignancy by transhiatal esophagectomy. Arch Surg 1988;123:1016-8. Sam HG, Pemberton JH, Payne WS. Management of instrumental perforations of the esophagus. J Thorac Cardiovasc Surg 1982;84:211-8. Orringer MB, Stirling MC. Esophagectomy for esophageal disruption. Ann Thorac Surg 1990;49:35-43. Skinner DB, Little AG, DeMeester TR. Management of esophageal perforation. Am J Surg 1980;139:760-4. Adkins MS, Raccuia JS, Acinapura AJ. Esophageal perforation in a patient with acquired immunodeficiency syndrome. Ann Thorac Surg 1990;50299-300. Bladergroen MR, Lowe JE, Postlethwait RW. Diagnosis and recommended management of esophageal perforation and rupture. Ann Thorac Surg 1986;42:235-9. Gouge TH, Depan HJ, Spencer FC. Experience with the Grill0 pleural wrap procedure in 18 patients with perforation of the thoracic esophagus. Ann Surg 1989;209:612-7. Attar S, Hankins JR, Suter CM, Coughlin TR, Sequeira A, McLaughlin JS. Esophageal perforation: a therapeutic challenge. Ann Thorac Surg 1990;50:45-51. Radmark T, Sandberg N, Pettersson G. Instrumental perforation of the oesophagus. A ten year study from two ENT clinics. J Laryngol Otol 1986;100461-5. Borgeskov S, Brynitz S, Siemenssen 0. Perforation of the esophagus. Experience from a department of thoracic surgery. Scand J Thorac Cardiovasc Surg 1984;1893-6. Banks JG, Bancewicz J. Perforation of the oesophagus: experience in a general hospital. Br J Surg 1981;68:5804. Nesbitt JC, Sawyers JL. Surgical management of esophageal perforation. Am Surg 1987;53:183-91. Finley RJ, Pearson FG, Weisel RD, Todd TRJ, Ilves R, Cooper J. The management of nonmalignant intrathoracic esophageal perforations. Ann Thorac Surg 1980;30575-83. Han SY, McElvein RB, AIdrete JS, Tishler JM. Perforation of the esophagus: correlation of site and cause with plain film findings. AJR 1985;145:537-40. Backer CL, LoCicero J 111, Hartz RS, Donaldson JS, Shields T. Computed tomography in patients with esophageal perforation. Chest 1990;98:1078-80. Wychulis AR, Fontana RS, Payne WS. Instrumental perforation of the esophagus. Chest 1969;55:18&9. Stewart ET, Miller WN, Hogan WJ, Dodds WJ. Desirability of roentgen esophageal examination immediately after pneumatic dilatation for achalasia. Radiology 1979;130: 589-91. Anderson RL. Spontaneous rupture of the esophagus. Am J Surg 1957;93:282-6. Groves LK. Instrumental perforation of the esophagus: what constitutes conservative management. J Thorac Cardiovasc Surg 1966;52:14. Nealon TF, Templeton JY, Cuddy VD, Gibbon JH. Instrumental perforation of the esophagus. J Thorac Cardiovasc Surg 1961;41:75-9. Paulson DL, Shaw RR, Kee JL. Recognition and treatment of esophageal perforation. Ann Surg 1960;152:13-6. Foley MJ, Ghahremani GG, Rogers LF. Reappraisal of contrast media used to detect upper gastrointestinal perforations: comparison of ionic water soluble media with barium sulfate. Radiology 1982;144:231-7. Phillips LG, Cumingham J. Esophageal perforation. Radio1 Clin North Am 1984;22:1607-13. Moghissi K, Pender D. Instrumental perforations of the oesophagus and their management. Thorax 1988;43:642-6. Santos GH, Frater RWM. Transesophageal irrigation for the

Ann Thorac Surg 1992;53534-43

treatment of mediastinitis produced by esophageal rupture. J Thorac Cardiovasc Surg 1986;91:57-62. 69. Endicott JN, Molony TB, Campbell G, Bartels LF. Esophageal perforations: the role of computerized tomography in diagnosis and management decisions. Laryngoscope 1986; 96:751-7. 70. Maroney TP, Ruiz EJ, Gordon RL, Pelligrini CA. Role of interventional radiology in the management of esophageal leaks. Radiology 1989;1701055-7. 71. Brewer LA, Carter R, Mulder GA, Stiles QR. Options in the management of perforations of the esophagus. Am J Surg 1986;152:62-9. 72. Richardson JD, Martin LF, Borzotta AP, Polk HC. Unifying concepts in the treatment of esophageal leaks. Am J Surg 1985;149:157-62. 73. DeMeester TR. Perforation of the esophagus [Editorial]. Ann Thorac Surg 1986;42231-2. 74. Menguy R. Near-total exclusion by cervical esophagostomy and tube gastrostomy in the management of massive esophageal perforation. Ann Surg 1971;173:613-6. 75. Gayet B, Breil P, Fekete F. Mechanical sutures in perforation of the thoracic esophagus as a safe procedure in patients seen late. Surg Gynecol Obstet 1991;172:125-8. 76. Engelberg M, Jedeikin RJ, Eschkol D, Hoffman S, Reiss R. Use of a stapling technique in closure of perforation of the esophagus. Am J Surg 1981;142:300-1. 77. Schorlemmer GR, Bums WW, Buckwalter JA. Extrapleural exclusion of esophageal perforations. South Med J 1984;77 1609-10. 78. Larsson S, Pettersson G. Advisability of concomitant immediate surgery for perforation and underlying disease of the esophagus. Scand J Thorac Cardiovasc Surg 1984;18:275-8. 79. McKinnon WM, Ochsner JL. Immediate closure and Heller procedure after Mosher bag rupture of the esophagus. Am J Surg 1974;127115-8. 80. Grillo HC, Wilkins EW Jr. Esophageal repair following late diagnosis of intrathoracic perforation. Ann Thorac Surg 1975;20:387-99. 81. Dooling JA, Zick HR. Closure of an esophagopleural fistula using onlay intercostal pedicle graft. Ann Thorac Surg 1967;3:55>7. 82. Bryant LR, Eiseman B. Experimental evaluation of intercostal pedicle grafts in esophageal repair. J Thorac Cardiovasc Surg 1965;50:626-9. 83. Millard AH. ‘Spontaneous’ perforation of the oesophagus treated by utilization of a pericardial flap. Br J Surg 1971;58: 70-2. 84. Hopper CL, Berk PD, Howes EL. Strength of esophageal anastomosis repaired with autogenous pericardial grafts. Surg Gynecol Obstet 1963;11783-6. 85. Rao KVS, Mir M, Cogbill CC. Management of perforation of the thoracic esophagus. Am J Surg 1974;12760%13. 86. Jara FM. Diaphragmatic pedicle flap for treatment of Boerhaave’s syndrome. J Thorac Cardiovasc Surg 1979;78:9314. 87. Thal AP, Hatafuku T. Improved operation for esophageal rupture. JAMA 1964;188:826-9. 88. Wise WS, Rivarola CH, Williams GD, Fink WJ,Read RC. Experience with the Thal gastroesophagoplasty. Ann Thorac Surg 1970;10:21>22. 89. Matthews HR, Mitchell IM, McGuigan JA. Emergency subtotal oesophagectomy. Br J Surg 1989;76918-20. 90. Miller RE, Tiszenkel HI. Esophageal perforation due to pneumatic dilation for achalasia. Surg Gynecol Obstet 1988; 16645840. 91. Tarkka M, Harkonen N, Lepojarvi M, Urkola P, Pokela R.

Ann Thorac Surg 19!32;5353443

92.

93. 94. 95. 96. 97. 98. 99.

100. 101. 102.

Repair of ruptured oesophagus. A case report. Acta Chir Scand 1985;151:385-7. Abbott OA, Mansour KA, Logan WD Jr, Hatcher CR Jr, Symbas PN. Atraumatic so-called "spontaneous" rupture of the esophagus. A review of 47 personal cases with comments on a new method of surgical therapy. J Thorac Cardiovasc Surg 1970;59:67-83. Sandrasagra FA, English TA, Milstein BB. The management and prognosis of oesophageal perforation. Br J Surg 1978; 65:62932. Asplund CM, Hill LD. Delayed lower esophageal perforation: management with Celestin tube. Ann Otol Rhino1 Laryngol 1985;94:114-6. Wesdorp IC, Bartelsman JF, Huibregtse K, den Hartog Jager FC, Tytgat GN. Treatment of instrumental oesophageal perforation. Gut 1984;25:398404. Imre J. Plastic tube prosthesis for the surgical treatment of perforations in esophageal strictures. Ann Thorac Surg 1973;15:275-80. Urschel HC, Razzuk MA, Wood RE, et al. Improved management of esophageal perforation: exclusion and diversion in continuity. Ann Surg 1974;179:587-92. Hausmaniger C, Roka R, Niederle 8. Technical modification of surgical therapy of esophageal perforation. Chirurg 1989; 60:55-7. Johnson J, Schwegman CW, Kirby CK. Esophageal exclusion for persistent fistula following spontaneous rupture of the esophagus. J Thorac Surg 1956;32:827-31. Sealy WC. Rupture of the esophagus. Am J Surg 1963;105: 505-10. Rosoff L, White EJ. Perforation of the esophagus. Am J Surg 1974;128207-18. Cameron JL, Kieffer RF, Hendrix TR, Mehigan DG, Baker RR. Selective nonoperative management of contained intrathoracic esophageal disruptions. Ann Thorac Surg 1979;27 4044.

103. Mollitt DL, Schulinger JN, Santulli TV. Selective management of iatrogenic esophageal perforation in the newborn. J Pediatr Surg 1981;16:989-93.

REVIEW JONES AND GINSBERG ESOPHAGEAL PERFORATION

543

104. Brown RH, Cohen PS. Nonsurgical management of spontaneous esophageal perforation. JAMA 1978;240140-2. 105. Lyons WS, Seremetis MG, deGuzman VC, Peabody JW Jr. Ruptures and perforations of the esophagus: the case for conservative supportive management. Ann Thorac Surg 1978;25:34&50. 106. Mengold LR, Klassen KP. Conservation management of esophageal perforation. Arch Surg 1965;91:23240. 107. Swedlund A, Traube M, Siskind BN, McCallum RW. Nonsurgical management of esophageal perforation from pneumatic dilatation in achalasia. Dig Dis Sci 1989;34379-84. 108. Vantrappen G, Hellemans J. Treatment of achalasia and related motor disorders. Gastroenterology 1980;79:144-54. 109. Slater G, Sicular AA. Esophageal perforations after forceful dilatation in achalasia. Ann Surg 1982;195:1864. 110. Berry BE, Ochsner JL. Perforation of the esophagus. A 30-year review. J Thorac Cardiovasc Surg 1973;65:1-7. 111. Ajalat GM, Mulder DG. Esophageal perforations, the need for an individualized approach. Arch Surg 1984;1191318-20. 112. Goldstein LA, Thompson WR. Esophageal perforations: a 15-year experience. Am J Surg 1982;143:495-503. 113. Flynn AE, Verrier ED, Way LW, Thomas AN, Pelligrini CA. Esophageal perforation. Arch Surg 1989;12412114. 114. Graeber GM, Niezgoda JA, Albus RA, et al. A comparison of patients with endoscopic esophageal perforations and patients with Boerhaave's syndrome. Chest 1987;92:995-8. 115. Triggiani E, Belsey R. Oesophageal trauma: incidence, diagnosis, and management. Thorax 1977;32:241-9. 116. Loop FD, Groves LK. Esophageal perforations. Ann Thorac Surg 1970;10:571-87. 117. Jemerin EE. Results of treatment of perforation of the esophagus. Ann Surg 1948;128971-5. 118. Saabye J, Nielson HO, Anderson K. Long term observation following perforation and rupture of the esophagus. Scand J Thorac Cardiovasc Surg 1988;22:79-80. 119. Erwall C, Ejerblad S, Lindholm CE, Aberg T. Perforation of the oesophagus. A comparison between surgical and conservative treatment. Acta Otolaryngol (Stockh) 1984;97: 1850-92.

Esophageal perforation: a continuing challenge.

Perforation of the esophagus remains a diagnostic and therapeutic challenge. Currently, the most common cause of perforation is instrumentation of the...
1MB Sizes 0 Downloads 0 Views