Volume 13, Issue 2 , Pages 138-146, Summer 2008
Esophageal Diversion
Article Outline
Esophageal perforation remains a diagnostic and therapeutic challenge, heavily taxing the thoracic surgeon's judgment and technical skills as well as institutional resources. Mortality remains high; a recent meta-analysis of 726 esophageal perforations occurring between 1990 and 2003 revealed a mortality rate of 18%.1 There are several variables well understood to be associated with outcome, including the interval from the event to the time of diagnosis, the mechanism of injury, the site of perforation, and the presence of underlying esophageal pathologic problems. Delay in therapy past 24 hours is often cited as a critical determinant in treatment choices and outcome as mortality rates have been shown to double beyond this critical window.1, 2, 3 The mortality of spontaneous perforation (Boerhaave's syndrome) remains exceedingly high (36%), almost double that of iatrogenic perforation (19%) and five-fold greater than that of traumatic perforation (7%). This fact is most likely due to the ambiguities of diagnosis and the resultant delays in care associated with spontaneous esophageal ruptures.1 Cervical perforations have a consistently lower mortality than thoracic or abdominal perforations, due to the body's ability to handle infection confined to the neck and the severe consequences of mediastinitis or uncontrolled pleural or peritoneal sepsis.1
The principles of treatment for esophageal perforation include cardiorespiratory resuscitation, cessation of the inflammatory stimulus by halting extraluminal contamination, control of infection, nutritional support, and eventual restoration of gastrointestinal continuity. The options for management of the perforation are quite broad and include expectant observation, endoscopic techniques to repair or otherwise control the site of perforation, surgical exploration with primary repair and drainage, esophageal resection with immediate reconstruction and, in extreme circumstances, esophageal exclusion and diversion. Nonoperative management may be attempted when the following conditions exist: (1) The diagnosis is made before significant extraluminal contamination; (2) A contained leak is present with free drainage of contrast back into esophageal lumen as assessed by contrast esophagography; (3) Absence of underlying esophageal pathologic problem that will impede healing (eg, malignancy, distal obstruction); (4) Clinical signs of sepsis are minimal or not present; (5) An experienced physician, knowledgeable in the management of esophageal perforations, is available.1, 4, 5, 6 When these criteria do not apply or a patient clinically deteriorates, surgical intervention is required. Cervical perforations are generally treated with either surgical drainage alone or primary repair and drainage, excluding cases of malignancy or distal obstruction. In the first 24 hours, uncontained intrathoracic and intra-abdominal perforations generally will require surgical exploration with debridement of devitalized tissue, wide local drainage, and primary repair, buttressed by a vascularized soft-tissue flap.
Beyond 24 hours, primary repair may be utilized, although the subsequent leak rates have been demonstrated to be significantly increased.7 In addition, these patients often present with more advanced septic physiology that may limit the extent of surgical intervention. Moreover, the prolonged exposure of the periesophageal tissues to intraluminal contents may preclude effective esophageal repair due to tissue friability. In this circumstance, the choices include esophageal exclusion/diversion and drainage alone or in combination with buttressed mucosal repair, esophageal resection and immediate reconstruction, or esophageal resection with diversion and delayed reconstruction. The choice of procedures is dependent on the clinical scenario as well as the surgeon's experience. Notably, in the presence of either malignancy or end-stage benign disease, resection is preferable. Due to the significant challenge of a second reconstructive procedure in this patient population as well as the morbidity and potential mortality associated with a later major reoperation, single-stage procedures have been proposed with varying success. Of course, the advisability of immediate reconstruction depends on the patient's clinical status and comorbidities. Considerable judgment must be exercised by the surgeon in deciding whether a patient is physiologically fit enough to tolerate the additional operative time and risk inherent in foregut reconstruction in the setting of sepsis. When operation is being undertaken for perforations over 24 hours in duration, rarely will the patient be suitable for immediate restoration of foregut continuity.
T-tube diversion is a promising means of creating a controlled fistula with outcomes comparable to those of perforations detected in the first 24 hours8, 9 but has also been met with strong criticism.10 Additional methods of esophageal exclusion in continuity with a side esophagostomy, based on the publication by Urschel and coworkers,11 have been proposed to create more easily reversible anatomy. Side esophagostomy, however, can be technically difficult to construct, particularly in the obese neck, and can result in incomplete diversion of esophageal contents, leading to continued mediastinal contamination. Additionally, this technique requires a cervical location of the esophagostomy, making subsequent appliance management potentially challenging. Due to the relative rarity of the clinical scenario and the persistently high morbidity and mortality, the authors favor end cervicothoracic esophagostomy with gastrostomy tube and jejunostomy tube placement as the most reliable and expeditious means of esophageal diversion. Although this strategy does commit the patient to a second surgery at a later date, typically a retrosternal reconstruction utilizing stomach, colon, or jejunum, this approach offers the most dependable means of esophageal diversion and, therefore, the best chance of successfully managing the initial physiologic insult.
Preoperative preparation needs to be expeditious. Resuscitation and appropriate antibiotic therapy should be initiated immediately and the patient emergently transported to the operating room. Arterial blood pressure monitoring and right-sided central venous access, via the internal jugular or subclavian veins, are generally established. Esophagoscopy utilizing minimal air insufflation may be performed intraoperatively to define the location and extent of injury in patients who are candidates for potential repair. Endoscopic visualization is especially useful for assessment of iatrogenic endoscopic injuries (eg, injuries sustained during transesophageal echocardiography) where the distance between the mucosal “entry” wound and the muscular “exit” defect may be significant.
Operative Technique
Figure 1.
(A) To perform a cervical esophagostomy, the patient is positioned supine with a pad placed transversely under the scapulae and the neck extended. The chest and neck are prepped into the field with adequate infraclavicular exposure for tunneling the esophagostomy. A left cervical approach is generally performed with an incision medial to the left sternocleidomastoid muscle. Patients requiring esophageal diversion generally require long-term ventilatory support and, thus, the surgeon should be contemplating a potential tracheostomy location when planning the cervical incision. At times we attempt a more “J”-shaped incision to maximize the distance between a potential tracheostomy and the cervicotomy. (B) The subcutaneous tissues and platysma are divided with cautery and the plane of dissection is carried medial to the left sternocleidomastoid muscle. The left middle thyroid vein, left inferior thyroid artery, and left omohyoid muscle are subsequently divided, exposing the trachea and esophagus. Care is taken to avoid metal retractors in the region of the tracheoesophageal groove to prevent recurrent laryngeal nerve injury. The cervical esophagus is then mobilized, first by sharply developing the plane between the esophagus and spine. SCM = sternocleidomastoid muscle; n = nerve; v = vein; m = muscle.
Figure 2.
The esophagus is then completely mobilized, taking care not to injure the recurrent laryngeal nerves running in the tracheoesophageal grooves. The periesophageal plane is then developed bluntly into the thoracic inlet to mobilize an ample portion of esophagus. Ideally, adequate esophageal length is mobilized to accommodate an infraclavicular location of the esophagostomy. Thoracic placement of the end esophagostomy facilitates placement of a functional ostomy appliance, which can be a significant challenge in the supraclavicular location. Additionally, maintenance of a potential tracheostomy in the future is dramatically easier. The esophagus is subsequently divided with an EndoGIA stapler and the wound is irrigated.
Figure 3.
The surgeon determines the site for the stoma based on the length of available esophagus. A 2-cm skin defect is created at the appropriate site and a subcutaneous tunnel is fashioned bluntly to that site using a Kelly clamp.
Figure 4.
The cervical esophagus is drawn through the cutaneous defect on the chest wall. The staple line is sharply excised and the esophagostomy is constructed using interrupted 3-0 polypropylene sutures placed through the skin and the full thickness of the esophagus.
Figure 5.
If resection has not been undertaken, distal exclusion must be accomplished as well. A midline laparotomy is performed and the viscera are packed away. The left triangular ligament of the liver is divided with electrocautery and the left lobe of the liver is retracted away from the hiatus with a padded retractor. The phrenoesophageal ligament is divided and the distal esophagus is dissected out circumferentially. The distal esophagus may then be divided with a surgical stapler.
If resection has been performed in conjunction with an end cervical esophagostomy, the diaphragmatic hiatus must be closed to prevent future herniation of abdominal contents into the chest.
If attempts are being made to leave the esophagus in continuity with a side esophagostomy, the distal esophagus may be occluded with either a single absorbable suture (eg, no. 1 Vicryl) or a linear surgical stapler without division. Although this can eliminate the need for a second reconstructive procedure, the authors do not advocate this approach because spontaneous recanalization can occur within 2 weeks of occlusion, leading to further mediastinal contamination in a recovering, critically ill patient.
Figure 6.
Distal diversion is accomplished with a gastrostomy tube placed strategically on a portion of the stomach that will later be excised when the gastric conduit is constructed for restoration of gastrointestinal continuity. Additionally, a feeding jejunostomy is placed for long-term enteral access. All wounds are irrigated and closed.
Conclusions
Esophageal perforation remains an infrequent, yet challenging, problem for the thoracic surgeon. Options for care are rapidly evolving and little data are available for rigorous scientific evaluation. Nonoperative management is an acceptable means of managing early, contained perforations in patients who are not clinically ill. Endoluminal prostheses are an evolving technology that offers promise for treating rapidly diagnosed perforations in stable patients, a situation which most frequently arises in the iatrogenic setting. It is vital that all thoracic surgeons acquire experience in deployment of these devices as the impact on the care of esophageal perforations in the future is likely to be significant.
When a patient presents late after an esophageal disruption with advanced septic physiology, esophageal diversion is necessary. For thoracic disruption, we generally prefer to perform a thoracotomy, whenever possible, to effectively debride and drain the contaminated space, assess the possibility of a buttressed repair, and perform resection when indicated. In extreme circumstances, a thoracotomy may not be possible and chest drainage can be obtained quickly with tube thoracostomy.
Special mention should be made of the unique set of problems facing the patient with a cervicothoracic esophagostomy. In addition to the tremendous critical care challenges of this patient population, specific problems related to the esophagostomy exist. Recurrent laryngeal nerve damage with resultant vocal cord and swallowing dysfunction may occur, requiring strict adherence to aspiration precautions, the investigation of vocal cord function, and the utilization of speech therapy resources. In addition, if aspiration occurs, one must suspect stomal stenosis, which may be addressed with Hagar dilators in the outpatient setting.12 Stomal stenosis due to esophageal ischemia may require surgical revision including repositioning the esophagostomy. Salivary losses of up to 1 L a day can cause significant dehydration as well as electrolyte imbalances including hypernatremia, hypokalemia, hypocalcemia, and metabolic acidosis. Vitamin B12 replacement should be initiated in patients who have required gastrectomy. Finally, the surgeon should play an integral role in the physical and nutritional rehabilitation of the patient in preparation for the restoration of gastrointestinal continuity.
References
- Evolving options in the management of esophageal perforation. Ann Thorac Surg. 2004;77:1475–1483
- . Diagnosis and management of esophageal perforations. Am Surg. 1992;58:112–119
- . Reinforced primary repair of thoracic esophageal perforation. Ann Thorac Surg. 1995;60:245–248
- Nonoperative management of esophageal perforations. Is it justified? Ann Surg. 1997;225:415–421
- . Esophageal perforation (A reassessment of the criteria for choosing medical or surgical therapy). Arch Int Med. 1992;152:757–761
- Selective nonoperative management of contained intrathoracic esophageal disruptions. Ann Thorac Surg. 1979;27:404–408
- Delayed primary repair of intrathoracic esophageal perforation: is it safe?. J Thorac Card Surg. 1996;111:114–121
- Modified T-tube repair of delayed esophageal perforation results in a low mortality rate similar to that seen with acute perforations. Ann Thorac Surg. 2007;83:1129–1133
- . Esophageal perforation: emphasis on management. Ann Thorac Surg. 1996;61:1447–1451
- . Invited commentary. Ann Thorac Surg. 2007;83:1133
- Improved management of esophageal perforation: exclusion and diversion in continuity. Ann Surg. 1974;179:587–591
- . Reversing esophageal discontinuity. Semin Thorac Cardiovasc Surg. 2007;19:47–55
PII: S1522-2942(08)00034-2
doi:10.1053/j.optechstcvs.2008.05.001
© 2008 Elsevier Inc. All rights reserved.
Volume 13, Issue 2 , Pages 138-146, Summer 2008
