| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2005;80:1185-1190
© 2005 The Society of Thoracic Surgeons
Division of Thoracic Surgery, Department of Cardiothoracic Surgery, Weill Medical College of Cornell University, New York, New York
Accepted for publication April 14, 2005.
* Address reprint requests to Dr Korst, Weill Medical College of Cornell University, Room M404, 525 East 68th St, New York, NY 10021 (Email: rjk2002{at}med.cornell.edu).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentOnline Discussion ForumReferences |
|---|
METHODS: A retrospective analysis of a prospectively collected esophageal carcinoma database (1988 to 2004) was performed at a single institution. Operative and pathologic data were collected, as well as details concerning the incidence, clinical features, treatment, and outcome of anastomotic leaks. Patients with leaks were further analyzed on the basis of whether or not intrathoracic manifestations of anastomotic leakage were present.
RESULTS: Two hundred, forty-two patients underwent transthoracic esophagectomy with a cervical anastomosis during the study period. There were 27 (11.1%) anastomotic leaks. Of these, 14 patients (52%) had intrathoracic manifestations of their cervical leaks, with empyema being the most common. Patients with intrathoracic spread of sepsis had significantly longer in-hospital (p < 0.001) and anastomotic healing times (p < 0.05) and required more drainage procedures (including reoperation; p < 0.005) than those with leaks confined to the neck. However, no difference in operative mortality or long-term survival was appreciated.
CONCLUSIONS: Intrathoracic manifestations of cervical anastomotic leaks are more common after transthoracic esophagectomy than what has historically been reported for transhiatal esophagectomy. This discrepancy may be due to anatomical or technical differences, or both, between the two procedures. Early diagnosis and aggressive drainage are necessary for achieving a favorable outcome.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentOnline Discussion ForumReferences |
|---|
A cervical anastomosis between the replacement conduit and the proximal esophageal stump is required during the performance of transhiatal esophagectomy (ie, esophagectomy without thoracotomy). Published literature has confirmed that the vast majority of anastomotic leaks after transhiatal esophagectomy are effectively treated by simply opening the cervical wound and initiating local wound care [4]. Due to the septic potential of anastomotic leakage, a cervical anastomosis is also routinely performed by many surgeons who practice transthoracic esophagectomy [5]. This is based on the assumption that if anastomotic failure should occur, any resulting infection would be limited to the neck. Despite this practice, there are little or no published data concerning the clinical manifestations and treatment of anastomotic leakage after transthoracic esophagectomy with a cervical anastomosis.
Transthoracic esophagectomy differs from the transhiatal approach in critical ways that may affect the manner in which cervical anastomotic disruptions present, such that drainage of the neck wound alone may not be sufficient treatment. First, during the transthoracic approach, the thoracic inlet tends to be much more widely dissected than in the transhiatal procedure. Second, after thoracotomy, the chest is drained using tubes, and suction is frequently applied. Given these differences, we hypothesized that the clinical manifestations and treatment of cervical anastomotic leakage after transthoracic esophagectomy may differ from that seen with the transhiatal approach. In this regard, the purpose of the present study is twofold: (1) to determine the incidence of cervical leaks that present with intrathoracic manifestations after esophagectomy with thoracotomy, and (2) to describe the treatment, clinical course, and outcome of these leaks.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentOnline Discussion ForumReferences |
|---|
The database includes specific fields for the type of esophagectomy performed, for the location of the anastomosis, and for the presence of anastomotic leak. Data entry for this database is performed by the thoracic data manager, based on data sheets completed by individual attending thoracic surgeons each time an esophagectomy is performed.
Data Collection and Definitions
Institutional review board approval was obtained for the purposes of data collection for this study. After identification of the study cohort, data were obtained from operative, pathology, and radiology reports, and progress notes and discharge summaries for all patients. All chart review and data collection was performed by a single investigator (RJK). Data collected included patient demographics, tumor characteristics, operative approach, and extent of lymphadenectomy, and the incidence, clinical features, treatment, and outcome of all anastomotic leaks. Associated complications as well as operative mortality and long-term survival were also recorded. The use of neoadjuvant therapy was not addressed in the present study due to a paucity of these cases. It is the practice of the authors to utilize neoadjuvant therapy only under the auspices of an approved protocol of the institutional review board. Therefore, the vast majority of patients underwent esophagectomy alone.
Tumor characteristics included histology, as well as tumor stage, according to the 1997 American Joint Commission on Cancer staging system. Transthoracic esophagectomy was subdivided into whether the thoracotomy was left or right. A two-field lymphadenectomy indicated that all lymph node-bearing tissue from the carina to the diaphragm was removed, as well as an upper abdominal and retroperitoneal node dissection to the celiac axis. A three-field procedure included the addition of the superior mediastinum and cervical region, specifically along both recurrent laryngeal nerves. Esophagogastrostomy was performed using either a single layer or double layer hand-sewn technique. Operative mortality was defined as either in-hospital death or death occurring within 30 days after the esophagectomy.
An anastomotic leak was defined prior to data collection and included both symptomatic as well as asymptomatic leaks. The authors evaluate all awake and lucid patients with a barium esophagram 6 to 7 days after resection, regardless of symptoms. Mechanically ventilated patients undergo esophagography once extubated and lucid. If an unexplained source of infection is believed to be present in an intubated patient, evaluation with chest computed tomography is performed, as well as bedside flexible esophagoscopy. Patients unable to safely perform esophagography (those with severe respiratory compromise such that aspiration could induce respiratory failure) are evaluated with either chest computed tomography or esophagography once their respiratory status has stabilized. Healing of anastomotic leaks was confirmed using either flexible esophagoscopy or contrast esophagram. Healing time was computed from the date of resection to the date of endoscopic or radiographic documentation of healing. Anastomotic dilatation was performed only for clinically significant dysphagia after healing of the anastomosis.
Intrathoracic manifestations of anastomotic leaks were defined prior to data collection and included foci of sepsis inferior to the thoracic inlet, including mediastinal abscess, empyema, conduit necrosis, massive intrathoracic hemorrhage, or tracheoesophageal fistula below the neck. Manifestations of anastomotic dehiscence were defined by both clinical signs and symptoms, as well as radiographic findings.
Data Analysis and Statistical Comparisons
Patients with anastomotic leakage were separated according to whether the manifestations of the leak were confined to the neck, or if they had spread to the thorax. Comparison of means was performed using independent samples of the Students t-test. Proportions were compared using chi2 contingency tables when expected frequencies were equal or greater than 5. When the expected frequencies were less than 5, Fisher's exact test was utilized. Survival was computed by the Kaplan-Meier method, with statistical significance determined by the log-rank method. Values of p less than 0.05 were considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentOnline Discussion ForumReferences |
|---|
|
|
Treatment of Anastomotic Dehiscence
In general, anastomotic dehiscence was treated immediately after diagnosis. The details of treatment are illustrated in Table 3. If a leak was confined to the neck and clinically occult, treatment consisted of antibiotics alone. However, all clinically overt leaks were addressed by establishing effective drainage. Whereby merely opening the cervical wound and providing local wound care sufficed for all patients with clinically overt leaks confined to the neck, approximately three times as many invasive procedures were required to establish drainage in the patients with intrathoracic complications, with several patients requiring more than one procedure. In this regard, nearly half the patients with intrathoracic sepsis needed to return to the operating room to effectively establish drainage, with a computed tomographic-guided drain playing a significant role as well. Despite this, no patient had an attempted anastomotic repair. However, 1 patient with a necrotic conduit was immediately reconstructed using a colon interposition.
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentOnline Discussion ForumReferences |
|---|
Intrathoracic Manifestations of Cervical Anastomotic Leaks are Common After Transthoracic Esophagectomy
Although clinical manifestations of the leaking cervical anastomosis are well-described after transhiatal esophagectomy, there is comparatively little published data that describes the immediate sequelae of cervical leaks after transthoracic esophagectomy. In perhaps the largest published series in which details are provided, Swanson and colleagues [5] reported an overall leak rate of 8% in a series of 250 consecutive patients who underwent transthoracic esophagectomy with cervical anastomosis. All patients with a simple anastomotic leak were effectively treated with local wound care; however, 26% of leaks were referred to as a conduit leak from the lesser curve staple line, reportedly not involving the anastomosis. Although not explicitly stated, it is implied that these latter patients required more extensive drainage or diversion. In the present series, slightly more than half of the patients with anastomotic dehiscence had intrathoracic manifestations of their leaks develop, which is in contrast to the previously mentioned report by Swanson and colleagues [5]. However, there were no patients with strictly a conduit leak from the lesser curve staple line; all staple line leaks were at the end of the conduit, were at least partially involved with the anastomosis, and seemed to be related to ischemia.
Further review of the published literature reveals additional, albeit smaller series of patients who underwent transthoracic esophagectomy with cervical anastomosis [8–10]. The results of the present study, in combination with these previously published reports, suggest that placement of the anastomosis in the neck during transthoracic esophagectomy does not eliminate the potential for intrathoracic complications.
Possible Mechanisms of Intrathoracic Extension of Sepsis From Cervical Leaks
Intrathoracic manifestations of cervical anastomotic dehiscence are not uncommon after transthoracic esophagectomy. Given that a cervically placed anastomosis is anatomically only 2 to 3 centimeters from the thoracic inlet, this is not surprising. However, intrathoracic extension of cervical leaks appears to be much less common after transhiatal esophagectomy, despite placement of the anastomosis in a similar location. Possible explanations for this discrepancy may include that the thoracic inlet is opened more widely during the transthoracic approach. During transhiatal esophagectomy, the pleural spaces are avoided, and the pleural envelope in the superior mediastinum remains intact, thus effectively protecting the pleural space from soilage as a consequence of a cervical leak. In contrast, during the transthoracic approach, the esophagus is dissected from the pleural space as well as the neck, establishing an open pathway between the cervical pre-vertebral and pleural spaces. In addition, many surgeons apply suction to the chest drains, which may facilitate leak extension into the chest. Finally, if one assumes that larger leaks take more time to heal than the small leaks, the observation in the present series that the leaks with intrathoracic manifestations took significantly longer to heal than those confined to the neck suggests that perhaps the former were larger leaks compared with the latter, and perhaps these were associated with more significant conduit ischemia. However, no objective characterization of leak size can be made from the present data.
Clinical Ramifications of Cervical Anastomotic Leaks with Extension into the Chest
Whatever the mechanism, intrathoracic manifestations of cervical leaks after transthoracic esophagectomy have serious consequences. Historically, mortality has been significant for these patients [5, 8–10]. However, in the present series, despite 14 patients with this complication, there was no 30-day or in-hospital mortality. In addition, and in contrast with a previously published large series of patients who underwent esophagogastrectomy, anastomotic leakage did not predispose to worsened long-term survival [11]. The lack of operative mortality in the present series may be attributable to early, aggressive investigation and drainage of leaks. In this regard, it is the authors' practice to obtain computed tomography of the chest for any patient after esophagectomy with unexplained fever or leukocytosis. Although some have advocated early attempts at anastomotic repair [12], the authors advocate only procedures to establish effective drainage of septic foci, allowing healing to occur by secondary intention.
The data presented herein suggest that even though operative mortality is not increased, intrathoracic extension of cervical anastomotic dehiscence is a serious complication. Length of hospital stay is significantly prolonged, and in many instances patients undergo multiple invasive drainage procedures. Despite these findings, the rate of additional, concomitant complications was nearly identical to that seen when leaks were confined to the neck. Once again, a policy of early intervention for any patient with a suspected leak perhaps may be credited and should be reinforced.
|
Online Discussion Forum |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentOnline Discussion ForumReferences |
|---|
Once the article selected for discussion is published in the online version of The Annals, we will post a notice on the CTSNet home page ( http://www.ctsnet.org ) with a FREE LINK to the full-text article. Readers wishing to comment can post their own commentary in the discussion forum for that article, which will be informally moderated by The Annals Internet Editor. We encourage all surgeons to participate in this interesting exchange and to avail themselves of the other valuable features of the CTSNet Discussion Forum and Web site.
For October, the article chosen for discussion under the Adult Cardiac Dilemma Section of the Discussion forum is:
Low Hematocrit During Cardiopulmonary Bypass is Associated With Increased Risk of Perioperative Stroke in Cardiac Surgery
Keyvan Karkouti, MD, George Djaiani, MD, Michael A. Borger, MD, PhD, William S. Beattie, MD, PhD, Ludwik Fedorko, MD, PhD, Duminda Wijeysundera, MD, Joan Ivanov, PhD, and Jacek Karski, MD
Tom R. Karl, MD
The Annals Internet Editor
UCSF Children's Hospital
Pediatric Cardiac Surgical Unit
505 Parnassus Ave, Room S-549
San Francisco, CA 94143-0118
Phone: (415) 476-3501
Fax: (212) 202-3622
e-mail: mailto:karlt{at}surgery.ucsf.edu
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentOnline Discussion ForumReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. T. Nguyen, M. W. Hinojosa, C. Fayad, and S. E. Wilson Minimally Invasive Management of Intrathoracic Leaks After Esophagogastrectomy Surgical Innovation, June 1, 2007; 14(2): 96 - 101. [Abstract] [PDF]
|
|
|
|
 |
|
 R. F. Heitmiller Intrathoracic manifestation of cervical anastomotic leaks. Ann. Thorac. Surg., July 1, 2006; 82(1): 383 - 383. [Full Text] [PDF]
|
|
|
|
|
|
R. J. Korst Reply. Ann. Thorac. Surg., July 1, 2006; 82(1): 383 - 384. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
