REVIEW
The Role of Esophageal Stents in the Management of Esophageal Anastomotic Leaks and Benign Esophageal Perforations Bobby V.M. Dasari, MRCS,∗ David Neely, MRCS,∗ Andrew Kennedy, FRCS,∗ Gary Spence, FRCS,† Paul Rice, FRCR,‡ Eamon Mackle, FRCS,‡ and Emmanuel Epanomeritakis, FRCS‡ Objective: The aim of this review was to assess the safety and effectiveness of esophageal stents in the management of benign esophageal perforation and in the management of esophageal anastomotic leaks. Background: Benign esophageal perforation and postoperative esophageal anastomotic leak are often encountered. Endoscopic placement of esophageal stent across the site of leakage might help control the sepsis and reduce the mortality and morbidity. Methods: All the published case series reporting the use of metallic and plastic stents in the management of postoperative anastomotic leaks, spontaneous esophageal perforations, and iatrogenic esophageal perforations were identified from MEDLINE, EMBASE, and PubMed (1990-2012). Primary outcomes assessed were technical success rates and complete healing rates. Secondary outcomes assessed were stent migration rates, stent perforation rates, duration of hospital stay, time to stent removal, and mortality rates. A pooled analysis was performed and subgroup analysis was performed for plastic versus metallic stents and anastomotic leaks versus perforations separately. Results: A total of 27 case series with 340 patients were included. Technical and clinical success rates of stenting were 91% and 81%, respectively. Stent migration rates were significantly higher with plastic stents than with metallic stents (40/148 vs 13/117 patients, respectively; P = 0.001). Patients with metallic stents had significantly higher incidence of postprocedure strictures (P = 0.006). However, patients with plastic stents needed significantly higher number of reinterventions (P = 0.005). Mean postprocedure hospital stay varied from 8 days to 51 days. There was no significant difference in the primary or secondary outcomes when stenting was performed for anastomotic leaks or perforations. Conclusions: Endoscopic management of esophageal anastomotic leaks and perforations with the use of esophageal stents is technically feasible. It seems to be safe and effective when performed along with mediastinal or pleural drainage. Esophageal stent can, therefore, be considered as a treatment option in the management of patients who present early after esophageal perforation or anastomotic leak with limited mediastinal or pleural contamination. Keywords: benign, esophagus, leaks, perforation, stents (Ann Surg 2014;259:852–860)
P
ostoperative anastomotic leak after esophageal surgery is encountered in 8% to 10% of patients undergoing esophageal resection.1,2 A Swedish national prospective study reported 8% anastomotic leak rates, with 20% associated mortality rate after intrathoracic resection and anastomosis for esophageal cancer.3 The From the ∗ Belfast Health and Social Care Trust, Belfast, United Kingdom; †Ulster Hospital, Belfast, United Kingdom; and ‡Craigavon Area Hospital, Craigavon, United Kingdom. Disclosure: The authors declare no conflicts of interest. Reprints: Bobby V. M. Dasari, MRCS, 15, The Blvd, Wellington Square, Belfast, BT7 3LW. E-mail: [email protected]. C 2014 by Lippincott Williams & Wilkins Copyright ISSN: 0003-4932/14/25905-0852 DOI: 10.1097/SLA.0000000000000564
852 | www.annalsofsurgery.com
England and Wales, esophagogastric cancer audit reported higher 30-day mortality rates in patients with anastomotic leak than those without leak (9% with leak vs 3.7% without leak; adjusted odds ratio: 2.82; 95% confidence interval: 1.59–5.01; P < 0.001).2 Early recognition and prompt management of leakage might help improve the outcome. Other causes of benign esophageal perforation include iatrogenic perforation after endoscopy, Boerhaave (spontaneous) perforation, and traumatic perforations due to foreign body or trauma. The principles of management of esophageal perforation and anastomotic leak include control of ongoing spillage from the esophagus, drainage of pleural and/or mediastinal cavities, intravenous broad-spectrum antibiotics, gastric decompression, and enteral or parenteral nutritional support. Sepsis is often controlled by surgical or percutaneous drainage of empyema, drainage of perianastomotic or mediastinal contamination, and with the use of intravenous antibiotics. The use of self-expanding stents in the palliation of symptomatic esophageal malignancies is well established and is effective.4 There is growing interest in the use of self-expanding stents in benign esophageal perforation. Endoscopic stenting is reported to be helpful to control sepsis and to allow early oral intake. However, placement of stent in a nonstrictured lumen could be associated with significant rates of stent migration, resulting in technical and clinical failure rates. In addition, endoscopic intervention for anastomotic leaks can be associated with further disruption of anastomosis, and other potential complications include perforation, bleeding, and strictures. We conducted a systematic review of the current literature to evaluate the success rates and safety of esophageal stents in the management of postoperative anastomotic leaks and benign esophageal perforations. Subgroup analysis was performed to assess the role of plastic and metallic stents and to assess the role of stents in patients with anastomotic leaks and perforations separately.
METHODS A systematic review of the literature was performed evaluating the role of esophageal stents in the management of postoperative anastomotic leaks and benign esophageal perforations.
Search Strategy and Study Selection A detailed electronic search was carried out from the following databases: MEDLINE, EMBASE, and PubMed. The search was performed using the Medical Subject Headings (MeSH): “esophagus,” “anastomotic leak,” OR “esophageal perforation,” “stents” combined with Boolean operator “AND.” Publications in English were included. All studies published from 1990 to 2012 were considered. Abstracts of potentially relevant publications based on the titles were reviewed, and case series reporting the outcomes of stent placement for postoperative anastomotic leaks and benign perforations were included. The reference list of all the included studies was hand-searched for further eligible studies that were then reviewed. Annals of Surgery r Volume 259, Number 5, May 2014
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 259, Number 5, May 2014
Authors of the case series that pooled the outcomes of patients with benign and malignant etiology were contacted by e-mail to request the data for patients with stenting for benign pathology separately. Where necessary, authors were contacted to clarify any additional information regarding follow-up duration, hospital stay, and postprocedure complications, if not explicit in the article. We identified 18 such articles and contacted the authors, of whom 4 responded with useful information that was included in the review. Study selection is presented as a flowchart in Figure 1.
Inclusion and Exclusion Criteria Prospective or retrospective case series with 3 or more than 3 patients in any single series evaluating the role of metallic or plastic esophageal stents in the management of esophagogastric anastomotic leaks, iatrogenic or traumatic esophageal perforations, and Boerhaave perforation were included.
Stents and Benign Esophageal Leak
Exclusion criteria applied were malignant perforations, anastomotic or benign strictures, stent for leakage from gastric staple line (sleeve gastrectomy), and gastrojejunal and colorectal anastomosis.
Data Extraction Two reviewers extracted all the data independently using paper data extraction pro forma. The information extracted from each study was year of publication, study design, inclusion criteria, exclusion criteria, number of participants in each study, type of stent used (metallic or plastic), number of stents used per patient, additional procedures performed, data on efficacy and safety of stents such as technical success rates, clinical success rates, failure rates, stent-related complication rates (stent migration, perforation, further dehiscence of anastomosis after stent insertion, bleeding), reintervention rates, mode of reintervention (endoscopic or surgical), and long-term complications such as anastomotic strictures and malignancy.
FIGURE 1. PRISMA flowchart illustrating search and data extraction history of articles on stenting of benign esophageal perforations and anastomotic leaks. C 2014 Lippincott Williams & Wilkins
www.annalsofsurgery.com | 853
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 259, Number 5, May 2014
Dasari et al
Primary outcomes evaluated were technical success rates and complete healing rates. Technical success was defined as being able to deploy 1 or more stents to occlude the site of leakage with no evidence of leakage of contrast at the end of the procedure. Clinical success was defined as complete healing of the perforation or leakage by placement of single or multiple stents irrespective of whether the stent was left in situ or was removed. The procedure was considered a clinical failure when the patient had to undergo surgical resection for persistent leak, died before confirming complete healing, or had a persistent fistula at the end of follow-up. Secondary outcomes were stent migration rates, reintervention rates, additional procedures performed, postprocedure perforation and bleeding, in-hospital mortality rates, time to stent removal, cost analysis, and long-term complications such as stricture or malignancy. Stent migration rates include the reported rates of stent dislodgement. Reintervention rates are those in which the patient needed repositioning of a stent (either because the stent had migrated or to control ongoing leakage), endoscopic placement of clips, or glue for partially healed leak and surgery for persistent leak. Removal of a dislocated stent by endoscopic or surgical means was not included in the reintervention rates. Additional procedures performed are defined as those performed to drain the sepsis or for providing nutritional support such as feeding jejunostomy or percutaneous endoscopic gastrostomy.
Subgroup analysis was performed to evaluate the role of metallic and plastic stents. Subgroup analysis was also performed in patient groups with anastomotic leaks and those with other causes of benign perforations. For subgroup analyses, studies reporting only the outcomes for the specific subgroups were included and those with mixed data (such as metallic and plastic stents) were not included.5,6
Statistical Analysis Pooled data analysis was performed to evaluate the primary and secondary outcomes. Where the data of certain outcomes are not provided, results are interpreted as a percentage of the number of patients from the included studies. Comparison of outcomes among subgroups was performed using 2 independent samples t test and test of equality of 2 proportions (SPSS and MINITAB).
RESULTS Twenty-seven studies were identified and included in the review. Study selection is presented as a flowchart illustrated in Figure 1. Details of the included studies and study characteristics are included in Tables 1 and 2. A total of 340 patients were included. Plastic stents were used in 12 studies that included 148 patients. Metallic stents were used in 12 studies that included 117 patients. Two studies with combined data using plastic and metallic stents in 70 patients5,6 and 1 study
TABLE 1. Details of the Included Studies Year
Type of Publication
Study Duration
Single Center
Total No. Patients
Cause of Esophageal Leak
Parajinen et Roy-Choudhury et al8 Siersema et al9 Evrard et al10 Gelbmann et al11 Hunerbein et al12 Koch et al13
1999 2001
Case series Case series
Not specified 1995–1999
Single center Single center
3 14
Iatro: 1, Boerhv: 1, Trauma: 1 Leaks: 14
2003 2004 2004 2004 2005
Case series Case series Case series Case series Case series
1999–2002 1999–2003 2002–2003 1998–2003 2001–2003
Single center Single center Single center Single center Single center
11 4 9 9 3
Langer et al14 Schubert et al15 Fischer et al16 Radecke et al17 Han et al18 Freeman et al19 Freeman et al20 Kim et al21
2005 2005 2006 2006 2006 2007 2007 2008
Case series Case series Case series Case series Case series Case series Case series Case series
2000–2003 2000–2003 1997–2004 2001–2005 2002–2005 Not specified Not specified 1998–2007
Single center Single center Single center Single center Multicenter Single center Multicenter Single center
24 12 15 9 8 17 21 17
Tuebergen et al22
2008
1999–2006
Single center
32
Iatro: 3, Leaks: 3, Boerhv: 5 Leaks: 4 Iatro: 3, Leaks: 5, Boerhv: 1 Leaks: 9 Iatro: 1, Leaks: 1, Postoperation fistula: 1 Leaks: 24 Leaks: 12 Spontaneous: 5, Iatro: 10 Iatro: 4, Leaks: 5 Leaks: 8 Iatro: 17 Leaks: 21 Iatro: 4, Leaks: 7, Fistulas: 2, Boerhv: 4 Leaks: 24, Iatro: 4, Boerhv: 4
Yano et al23 Amrani et al24 Babor et al25 Freeman et al26 Salminen et al27 Fernandez et al28 Kiernan et al29 van Heel et al6
2008 2009 2009 2009 2009 2010 2010 2010
Not specified 2005–2008 2004–2006 Not specified 1999–2008 2007–2008 1998–2009 2001–2009
Single center Single center Single center Single center Single center Single center Single center Single center
3 4 5 19 10 4 8 33
D’Cunha et al5 Cerna et al30 Inbar et al31
2011 2011 2011
2005–2009 2008–2010 2009–2010
Single center Single center Single center
37 5 4
Author al7
Nonrandomized observational study Case series Case series Case series Case series Case series Case series Case series Case series Case series Case series Case series
Iatro: 1, Leaks: 2 Perf: 2, Leaks: 2 Leaks: 4, Boerhv: 1 Spontaneous perfs: 19 Perfs: 8, Leak: 2 Leaks: 4 Perfs: 8 Iatrogenic: 19, Boerhv: 10, FB: 1, Trauma: 1, Leak: 1 Leaks: 22, Perfs: 15 Leaks: 4, Perfs: 1 Leaks: 3, Esophagopleural fistula: 1 (excluded patients with leak from sleeve gastrectomy)
Boerhv indicates Boerhaave; Iatro, iatrogenic perforations; Leaks, anastomotic leaks; Perfs, perforations.
854 | www.annalsofsurgery.com
C 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 259, Number 5, May 2014
Stents and Benign Esophageal Leak
TABLE 2. Characteristics of the Included Studies Metallic/ Plastic
Investigations to Confirm
Evrard et Fernandez et al28
P P
OGD, CS NM
Polyflex Polyflex
Freeman et al19
P
CS
Polyflex
Freeman et al20
P
CS
Polyflex
Freeman et al26
P
CS
Polyflex
Gelbmann et al11 Hunerbein et al12 Kim et al21
P P P
NM CS CS, CT
Langer et al14 Radecke et al17 Schubert et al15 Yano et al23 Amrani et al24 Babor et al25 Fischer et al16 Han et al18 Inbar et al31 Kiernan et al29 Koch et al13
P P P P M M M M M M M
NM NM OGD, CS CS NM NM OGD, CS, CT OGD, CS, CT OGD NM NM
Parajinen et al7 Roy-Choudhury et al8 Salminen et al27 Siersema et al9
M M
OGD, CS, CT CS
M M
OGD, CS, CT NM
Tuebergen et al22
Author al10
M
OGD
D’Cunha et al5 van Heel et al6
M, P M, P
OGD, CS, CT NM
Cerna et al30
BD
CS
Additional Procedures Performed
Type
Polyflex Polyflex Montgomery salivary bypass stents Polyflex Polyflex Polyflex Polyflex Niti-S: 3, Hanaro: 1 Covered Ella Boubella Ultraflex: 13, Niti-S: 2 Mushrom-shaped stents Sx-Ella, covered Covered wall stent Covered Dostent: 1, Hanaro: 1, Rosch: 1 Coated Schneider stent: 3 Ultraflex: 10, Gianturco: 3, Telestep: 4 Ultraflex: 1, Hanaro: 9 Wall stent: 7, Ultraflex: 4 Ultraflex Covered Ninitol/Polyflex Fer X Ella: 1, Flamingo: 3, Ultraflex: 35, Polyflex: 5, Hanarostent: 6 Ella–BD
Follow-up Period
NM Perianastomotic drainage: 2, tube thoracotomy: 2, radiological drain: 1 PEG: 12, esophageal dilatation: 1, thoracotomy: 4, VATS decortication: 2, celiotomy, gastrostomy, jejunostomy: 1 Tube jejunostomy: 10, thoracotomy: 5, PEG: 4, decortication: 4, celiotomy: 2, neck exploration: 1 PEG: 19, VAT decortication: 5, tube thoracotomy: 4 Chest drain in all: 9 Pleural drainage in all: 9 NM
NM Mean: 12 mo NM
NM Pleural/mediastinal drainage: 6 Pleural drainage: 14 Mediastinal drain: 1, gastrostomy: 1 None Surgical drainage: 5 Pleural drainage: 7, surgical: 2 Pleural drain: 8, jejunostomy: 3 None Chest drain: 1 Perianastomotic drainage: 1
Median: 220 d 40–565 d 9 mo 6–18 mo NM 4–12 mo 6 mo to 6 yr 7–30 mo NM Incomplete data 1–2 yr
Thoracotomy for drainage: 1 Pleural drainage: 3
Incomplete data 8 patient-months (range: 3–18 patient-months) NM Incomplete data
Pleural drainage: 4 Enteral feeding 2 patients, parenteral feeding 12 patients, pleural drainage 3 patients Pleural and perianastomotic drainage: 32 Drainage procedures: 19 Pleural drainage in all: 33
Gastrostomy: 1
8 mo NM >1 mo
>1 mo >3 mo
24 mo 41.6 mo Median: 2 yr
1–24 mo
BD indicates biodegradable; CS, contrast study; CT, contrast tomogram; M, metallic; NM, not mentioned; OGD, gastroscopy; P, plastic; PEG, percutaneous endoscopic gastrostomy; VATS, video-assisted thoracoscopic surgery.
reporting the use of biodegradable stents30 in 5 patients were included. Polyflex stents were the most often used plastic stents. Kim et al21 used Montgomery salivary bypass stents. Metallic stents were used in the remaining series. Subgroup analysis was performed to compare the role of plastic and metallic stents. In view of the limited numbers in the subgroups, we have not analyzed the covered and uncovered metallic stents. Nine case series with 100 patients reported use of stents for anastomotic leaks only; 4 case series with 54 patients reported use of stents for benign perforations only. Subgroup analysis was performed to compare the role of stents in these 2 groups. Diagnosis of perforation or anastomotic leak was made using contrast study, computed tomography, or gastroscopy. Contrast study was the commonly used mode of investigation whereas most studies used more than 1 modality of investigation. There was a wide variation C 2014 Lippincott Williams & Wilkins
in the duration of follow-up period among the included studies, from 1 month to 42 months (Table 2).
Primary Outcomes The primary outcomes evaluated were the technical and clinical success rates of esophageal stenting in benign perforation and anastomotic leakage.
Technical and Clinical Success Rates Technical success was achieved in 311 of 340 (91.4%) patients. Clinical success was achieved in 250 of 308 (81.1%) patients. Two studies14,29 did not report the complete healing rates, and these were excluded from the evaluation of clinical success rates. Patients with persistent leak or fistula (29 patients), those who died before confirmation of healing, and those lost to follow up (19 patients) from www.annalsofsurgery.com | 855
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 259, Number 5, May 2014
Dasari et al
the available data were considered as clinical failures to heal. Reason for clinical failure is not clear in 10 patients. Technical success rates in the plastic and metallic stent groups were 89.8% versus 96.5% (P = 0.025). Clinical success rates were similar in both the groups (86.2%). A total of 404 stents were used in 300 patients (1.34 stents per patient). A total of 174 plastic stents were used for 148 patients (1.17 plastic stents per patient), 83 metallic stents were used for 77 patients (1.07 metallic stents per patient), and 7 biodegradable stents were used for 5 patients (1.4 per patient). Two studies5,6 provided combined data of plastic and metallic stents. It was unclear from 2 studies about how many stents were used (40 patients).22,29
Secondary Outcomes
to healing is assessed after removal of stent, and it has, therefore, varied significantly between the studies.
Mortality Rates Forty-four deaths were reported, and it was not clear from the included studies whether the mortality was in-hospital, 30-day, or 90-day mortality rates. Also, not all of the deaths were as a result of perforation or sepsis related to perforation (Table 3).
Long-term Complications Eleven postprocedure strictures were reported in 340 patients (1 in plastic stents vs 9 in metallic stents; P = 0.006). No malignancies were reported in any of the series at follow-up.
Stent Migration
Leaks Versus Perforations
The reported overall stent migration rates were 73 of 340 patients (20.8%), 40 of 148 patients (27%) with plastic stents, and 13 of 117 patients (11%) with metallic stents (P = 0.001) (Table 3).
This subgroup analysis was performed from the studies that have reported these results separately (Table 4). There were no significant differences in the primary or secondary outcomes evaluated.
Stent Perforation Stent perforation was reported in 7 of 340 patients (2%) (plastic vs metallic: 2 vs 5; P = 0.69) (Table 3). Two patients were reported to have suffered from bleeding because of erosion of the stent into the aorta. Further dehiscence of the anastomosis was reported in 1 patient.
Stent-Related Bleeding Three patients were reported to have suffered bleeding after stent deployment.
Reintervention Rates Endoscopic reinterventions were reported in a total of 58 patients (17%) and surgical reintervention was reported in 33 patients (9.7%). Where the data were provided separately for the plastic and metallic stents, overall reintervention rates were significantly higher in the plastic stent group than in the metallic stent group (41 vs 13; P = 0.005). Endoscopic reinterventions were required in 22% versus 5% and surgical intervention was required in 5.8% versus 15% between the plastic and metallic groups (P < 0.00001 and 0.14, respectively).
Additional Procedures Performed The role of esophageal stenting in the management of patients with esophageal perforation or anastomotic leaks is to limit the sepsis from continued leak and to allow early resumption of enteral feeding to attend the nutritional needs. However, most of these patients need additional procedures for the management of established or coexisting sepsis in the pleural cavity or mediastinum. Thoracotomy (tube, surgical or video-assisted) was the most common additional procedure performed, reported in 187 patients. The use of percutaneous gastrostomy (35 patients), surgical gastrostomy (2 patients), jejunostomy (13 patients), and parenteral nutrition (12 patients) for nutritional support has been reported (Table 2).
Duration of Hospital Stay Duration of hospital stay was reported in 13 of the included studies. It varied from mean postoperative hospital stay of 8 days to 51 days, reflecting the complex and prolonged course of recovery.
Time to Stent Removal and Time to Healing There is significant variability in the reported studies regarding the time to stent removal (Table 3). Although some centers removed the stents at 4 weeks, others have left the stent for longer period. Time 856 | www.annalsofsurgery.com
DISCUSSION Twenty-seven case series reported the use of stents in the management of esophageal anastomotic leaks or benign esophageal perforations, and these studies represent the current available evidence on this subject. Currently, there are no published studies comparing the use of stents versus surgical or conservative management, nor comparing the role of metallic versus plastic stents for this group of patients. van Boeckel et al32 in their systematic review included patients with anastomotic dehiscence after bariatric surgery in addition to the patient groups included in our review and the data published until 2010. Patients who underwent bariatric surgery are not included in our review. All the remaining articles (an overlap of 18 studies between the reviews), other relevant studies (3), and the larger series published up to 2012 (6 additional studies) are included. Appropriate preprocedure workup of esophageal perforation typically includes a contrast study. Computed tomography allows evaluation of the severity of mediastinal or pleural space contamination. Endoscopic visualization helps identify the exact location of the perforation, extent of dehiscence, and the viability of esophageal mucosa. Investigations used in the preoperative workup among the included studies are listed in Table 2. Contrast study is the commonly used mode of investigation whereas most studies used more than 1 modality of investigation. The overall technical and clinical success rates of esophageal stent placement in these patient groups were 91% and 81%, respectively, similar to previously published data.32 The procedure is technically feasible, with most individual case series reporting 100% technical success rates. It is important to know whether the technical success is being translated to clinical success (complete healing) and the overall desired outcomes of the procedure (such as time to complete healing and duration of hospital stay). The variables that determine overall success might include complete healing of the leak, control of sepsis after stent deployment, lack of or limited reintervention rates, early return to oral intake, and low overall morbidity and mortality rates. Complete healing rates of the leak are certainly impressive (81%), and these are comparable with previously published reviews. There is no statistically significant difference in the complete healing rates between the plastic and metallic stents (P = 0.99). The most common source of sepsis in this patient group is pleural, mediastinal, or perianastomotic contamination, and 63% of patients required a form of drainage procedure for controlling sepsis. Severity of sepsis after esophageal perforation might also be determined by the delay associated with identification of leak. Fischer C 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
C 2014 Lippincott Williams & Wilkins
P M
M M
M
Yano et al23 Amrani et al24
Babor et al25 Fischer et al16
Han et al18
4 8
8
5 15
3 4
12
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Salminen et
M
10
P
Schubert et al15
9
14
P
Radecke et al17
24
al27
P
Langer et al14
9 17
Roy-Choudhury et al8 M
P P
Hunerbein et al12 Kim et al21
9
3 3
P
Gelbmann et al11
19
M M
P
Freeman et al26
21
Koch et al13 Parajinen et al7
P
Freeman et al20
17
M M
P
Freeman et al19
4 4
Inbar et al31 Kiernan et al29
P P
Fernandez et al28
Evrard et al10
Author
5
1
0 0
0 2
1
0 2
0 0
0
4
6
0 1
3
0
1
0
0 1
0
0 0
0 0
0
0
2
0 0
0
0
0
0
0 0
1
0 0
0
2 0 1 1 (eroded into thoracic aorta) 3 0 0 0
0
2 0
1 0
2
2
9
2 6
3
4
5
3
3 0
0
1
0 0
0 0
0
0 0
0 0
0
0
0
0 0
0
0
0
0
0 0
Restent: 2
Restent: 2, repositioning: 4
Restent/ repositioning: 4 Restent: 1, repositioning: 3 Restent: 2 Repositioning: 4
Restent/ repositioning: 3 Restent/ repositioning: 5
2
0
2 0
1 4
0
0 0
1 1
0 0
Repositioned: 1
0
1 0
Repositioned: 1 Unclear
0
0 Restent + clips + glue: 1 0 0
1
0
1 0
0 1
0
0 0
1 0
0
0
3
0 0
0
2
1
1
0 0
1
0
1 1
0 0
1
0 2
0 0
0
0
0
0 0
0
0
1
0
0 0
0
0
0 0
0 0
0
0 0
0 0
0
0
0
0 0
0
0
0
0
0 0
13
14
4 3
4 Unclear
8
5 15
3 5
12
11
27
18 17
10
21
27
19
4 5
8
14
3 3
4 7
8
5 15
3 4
12
9
16
8 17
7
17
20
16
4 4
Reason for Clinical Failure
Time to Stent Removal
... Not mentioned 2 mortality, 1 lost 4 wk to follow up—so clinical success could not be confirmed 8 ... 18–48 d; mean 32 ± 11 d 4 ... 6–7 wk Unclear Outcome after 4–6 wk stent removal is not mentioned 3 ... Unclear 3 ... Not removed the stents 13 Persistent leak and Unclear death: 1 8 Persistent leak and 3–16 wk; death: 1, persistent median: 10 wk leak requiring surgical resection: 1 (continued)
5 12
... 1 wk to 4 mo Early mortality: 1, 4–8 wk minor persistent fistula: 1 16 Continued leak req Range: 17–91 d; surgery: 1 52 ± 20 d 20 Complete Range: 15–175 d; dehiscence of mean: 51 ± 43 d existing surgical repair due to stent placement: 1 17 Persistent leak: 2 Range: 14–32 d; mean: 20 ± 15 d 6 Death: 3 Range: 32–242 d; mean: 135 ± 78 d 9 ... mean: 29 d 13 Persistent fistula: 3 Range: 12-109d; mean: 38 ± 25 d Unclear 4 were not assessed Unclear by postcontrast study 6 Death: 3 Range: 15–438 d; median: 55.5 d 11 Persistent leak: 1 Range: 2–8 wk; median: 4 wk 3 ... 5–10 wk 4 ... Average: 5 wk
4 2
Surgery as Total Restent/ Reinterven- Strict- Malign- No. Technical Clinical Repositioning tion ure ancy Stents Success Success
UnclearUnclear
2
9
2 4
4
6
8
5
Unclear 0
Total Number Metallic/ No. of Migra- Number of ReinterPlastic Patients Mortality tions Perforations Bleeding ventions
TABLE 3. Primary and Secondary Outcomes of Patients Managed With Plastic (P) Stents and Metallic (M) Stents∗
Annals of Surgery r Volume 259, Number 5, May 2014 Stents and Benign Esophageal Leak
www.annalsofsurgery.com | 857
Annals of Surgery r Volume 259, Number 5, May 2014
5 7 0 0 0 1 1 0 0 3 0 5 BD
Two studies reported combined data with both (B) plastic and metallic stents and 1 with biodegradable (BD) stents. BD indicates biodegradable; M, metallic; P, plastic.
Cerna et al30
∗
11 B van Heel et al6
858 | www.annalsofsurgery.com
33
37 B D’Cunha et al5
7
4
4 required surgery 5 wk (R: 1–84 wk) for stent failure, 7 deaths (total 10) Persistent leak: 1 ... 23 32 46 0 1 4 Repositioning: 4, restent: 4 12 2
22 28 94 0 0 9 16 25 0
3 (gastroaortic fistula: 1) 1 6
0 0 3 5 32 Tuebergen et al22 M
5
8
3
0
Unclear
32
25
Technical failure 1, Median: 7 wk sepsis due to (R: 6–14 wk) abscess 1: both requiring esophagectomy Unclear Median: 46 d (R: 4–426 d) Nonseal: 9 patients Not mentioned 9 10 12 0 0 2
Endoscopy and removal of metallic strand causing dysphagia: 1 UnclearUnclear 3 0 0 1 0 11 M Siersema et al9
Author
Total Number Metallic/ No. of Migra- Number of ReinterPlastic Patients Mortality tions Perforations Bleeding ventions
TABLE 3. (Continued)
Surgery as Restent/ Reinterven- StrictRepositioning tion ure
Total Malign- No. Technical Clinical ancy Stents Success Success
Reason for Clinical Failure
Time to Stent Removal
Dasari et al
et al16 reported that stent insertion within 1 hour resulted in uneventful recovery, with average postoperative hospital stay of 5 days compared with patients who had a stent inserted with a median delay of 123 hours. In the delayed group, the clinical course was complicated with multiorgan failure, and the average postoperative hospital stay was 44 days.16 None of the included studies reported the severity of sepsis at the time of endoscopic intervention. Successful deployment of stent controlling the leakage or spillage should ultimately result in complete healing provided the stent remains in the required position until the dehiscence heals. Yakoub et al4 reported migration rates of 5% versus 10% (P = 0.09) in their meta-analysis comparing the use of metallic and plastic stents for patients with malignant dysphagia. With no stricture and hinge to keep the stent in place, the migration rates are expected to be higher with its use in benign nonstricturing etiology. Overall stent migration rates in our review are 20.8% and are significantly better with the metallic stents (P = 0.001). Despite the reported high migration rates, clinical healing rates are, however, similar between the 2 groups. This could be either due to migration of the stent after complete healing or due to reintervention undertaken after migration of the stent. The most common endoscopic reintervention is placement of another stent due to persistent leak, stent migration, or repositioning of the stent. Endoscopic placement of fistula plugs or resolution clips for persistent fistula is also reported. From this review, 13% mortality rates with the endoscopic management of perforations and leaks seem to be more favorable than the mortality rates reported with surgical intervention. However, we are not able to ascertain whether these are 30-day, 90-day, or even long-term mortality rates from the reported studies. From the available studies, it is difficult to evaluate the time taken for the perforation or leak to heal. Time to heal can be accurately ascertained only after removing the stent, followed up by a contrast study or direct visualization. Difficulty to remove stents after resolution of the leak/perforation due to hyperplasia of the tissue in the uncovered parts at both ends is reported and can induce stenosis that is a concern. Animal studies have suggested that 4 weeks should be sufficient time for tissue healing. Although the more recent studies have opted to remove the stent after 6 to 8 weeks,5,15,23,24,27,28,31 some of the earlier studies left the stent in situ much longer or even permanently. Another reason for ambiguity regarding the time to healing is whether the stent removal has caused disruption of healed mucosa, resulting in the leak of contrast during check contrast study. Cost considerations are also important when evaluating the 2 different intervention strategies. None of the included case series performed cost-effectiveness analysis. In an attempt to reduce the heterogeneity, we have included only those patients with anastomotic leaks and benign perforations. However, heterogeneity still remains a drawback of this current review, as it includes varied patient groups such as those with iatrogenic perforations after dilatation of a benign stricture and anastomotic leak after cancer resection. Outcomes of the patients with significant mediastinal or pleural contamination are expected to be worse than those with minimal contamination and lack of this differentiation adds to the heterogeneity of the patient groups included in the study. Although this review is based on case series with small patient numbers, it provides the best available evidence on this topic and suggests that endoscopic stent placement is a safe, acceptable and effective minimally invasive procedure to manage patients with benign esophageal leaks. The review, however, does not provide comparative data between the standard surgical management and the novel endoscopic management and covered or uncovered stents. The role of stent in patients with significant mediastinal or pleural contamination and the time to removal of stent remain unclear. There is no evidence to suggest that stent placement is superior to other forms of standard C 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
C 2014 Lippincott Williams & Wilkins
P
P P
P
Freeman et al26
Fischer et al16 Freeman et al19
Parajinen et al7
L indicates leak; P, perforation.
L
Inbar et al31
3 54
15 17
4 100 19
4 4 8 9 12 24
21
L
L L L L L L
14
L
0 2
2 0
0 10 0
0 1 1 0 0 6
1
1
0 7
0 3
2 23 4
3 0 0 2 2 9
5
0
0 0
0 0
0 2 0
0 0 0 0 0 2
0
0
0 11
0 5
1 20 6
Unclear 0 0 2 Unclear 9
8
0
Leak/ Total No. Mortality Stent Migration Stent Perforation Reintervention Perforation Patients Rates Rates Rates Rates
Evrard et Fernandez et al28 Han et al18 Hunerbein et al12 Schubert et al15 Langer et al14
al10
Roy-Choudhury et al8 Freeman et al20
Author
Restent/ repositioning: 4 0 Restent/ repositioning: 3 0
Restent/ repositioning: 5 0 0 0 Restent: 2 Unclear Restent: 2, repositioning: 4 Repositioned: 1
0
Restent
0
0 1
2
0
0 0 0 0 0 3
1
0
1 3
2 0
0 2 0
0 0 1 0 0 0
1
0
3 58
15 19
4 119 21
4 5 8 18 12 27
27
14
3 51
15 16
4 90 17
4 4 8 8 12 16
20
14
3 48
12 16
4 71 17
4 2 8 9 11 Unclear
20
13
Surgery as Total No. Technical Clinical Reintervention Stricture Stents Success Success
TABLE 4. Subgroup Analysis Comparing the Outcomes in the Anastomotic Leak Group Versus Perforations Group
Annals of Surgery r Volume 259, Number 5, May 2014 Stents and Benign Esophageal Leak
www.annalsofsurgery.com | 859
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 259, Number 5, May 2014
Dasari et al
treatment due to the lack of randomized controlled trials. There is a role to perform randomized controlled trials in homogeneous patient groups comparing the surgical and endoscopic management and various types of stents with well-defined outcomes including clinical success rates and to assess the optimum time for stent removal.
CONCLUSIONS Endoscopic management of esophageal anastomotic leaks and benign perforations is technically feasible. It seems to be a safe and effective management strategy when performed along with an additional drainage procedure to control sepsis. Esophageal stent can, therefore, be considered as a treatment option in the management of patients who present early after esophageal perforation or anastomotic leak with limited mediastinal or pleural contamination.
REFERENCES 1. Viklund P, Linblad M, Lu M, et al. Risk factors for complications after esophageal cancer resection. A prospective population-based study in Sweden. Ann Surg. 2006;243:204–211. 2. http://www.augis.org/pdf/NHS-IC-OGC-Audit-2010-interactive.pdf. Accessed January 31, 2013. 3. Rutegard M, Lagergren P, Rouvelas I, et al. Intrathoracic anastomotic leakage and mortality after esophageal cancer resection: a population-based study. Ann Surg Oncol. 2012;19:99–103. 4. Yakoub D, Fahmy R, Athanasiou T, et al. Evidence-based choice of esophageal stent for the palliative management of malignant dysphagia. World J Surg. 2008;32:1996–2009. 5. D’Cunha J, Rueth NM, Groth SS, et al. Esophageal stents for anastomotic leaks and perforations. J Thorac Cardiovasc Surg. 2011;142:39–46. 6. van Heel NC, Haringsma J, Spaander MC, et al. Short-term esophageal stenting in the management of benign perforations. Am J Gastroenterol. 2010;105:1515–1520. 7. Parajinen J, Ristkari SK, Mokka REM. A report of three cases with an oesophageal perforation treated with a coated self-expanding stent. Ann Chir Gynaecol. 1999;88:332–334. 8. Roy-Choudhury SH, Nicholson AA, Wedgwood KR, et al. Symptomatic malignant gastroesophageal anastomotic leak: management with covered metallic esophageal stents. Am J Roentgenol. 2001;176:161–165. 9. Siersema PD, Homs MYV, Haringsma J, et al. Use of large diameter metallic stents to seal traumatic non-malignant perforations of the esophagus. Gastrointest Endosc. 2003;58:356–361. 10. Evrard S, Le Moine O, Lazaraki G, et al. Self-expanding plastic stents for benign esophageal lesions. Gastrointest Endosc. 2004;60:894–900. 11. Gelbmann CM, Ratiu NL, Rath HC, et al. Use of self-expandable plastic stents for the treatment of esophageal perforations and symptomatic anastomotic leaks. Endoscopy. 2004;36:695–699. 12. H¨unerbein M, Stroszczynski C, Moesta KT, et al. Treatment of thoracic anastomotic leaks after esophagectomy with self-expanding plastic stents. Ann Surg. 2004;240:801–807. 13. Koch S, Weber A, Fein F, et al. Esophageal stents as a salvage therapy for non-malignant iatrogenic esophageal perforations. Gastroenterol Clin Biol. 2005;29:735–739. 14. Langer FB, Wenzl E, Prager G, et al. Management of postoperative esophageal leaks with the Polyflex self-expanding covered plastic stent. Ann Thorac Surg. 2005;79:398–403.
860 | www.annalsofsurgery.com
15. Schubert D, Scheidbach H, Kuhn R, et al. Endoscopic treatment of thoracic esophageal anastomotic leaks by using silicone-covered, self-expanding polyester stents. Gastrointest Endosc. 2005;61:891–896. 16. Fischer A, Thomusch O, Benz S, et al. Nonoperative treatment of 15 benign esophageal perforations with self-expandable covered metal stents. Ann Thorac Surg. 2006;81:467–472. 17. Radecke K, Lang H, Frilling A, et al. Successful sealing of benign esophageal leaks after temporary placement of a self-expanding plastic stent without fluoroscopic guidance. Z Gastroenterol. 2006;44:1031– 1038. 18. Han XW, Li YD, Wu G, et al. New covered mushroom-shaped metallic stent for managing anastomotic leak after esophagogastrostomy with a wide gastric tube. Ann Thorac Surg. 2006;82:702–706. 19. Freeman RK, Van Woerkom JM, Ascioti AJ. Esophageal stent placement for the treatment of iatrogenic intrathoracic esophageal perforation. Ann Thorac Surg. 2007;83:2003–2007. 20. Freeman RK, Ascioti AJ, Wozniak TC. Postoperative esophageal leak management with the Polyflex esophageal stent. J Thorac Cardiovasc Surg. 2007; 133:333–338. 21. Kim AW, Liptay MJ, Snow N, et al. Utility of silicone esophageal bypass stents in the management of delayed complex esophageal disruptions. Ann Thorac Surg. 2008;85:1962–1967. 22. Tuebergen D, Rijcken E, Mennigen R, et al. Treatment of thoracic esophageal anastomotic leaks and esophageal perforations with endoluminal stents: efficacy and current limitations. J Gastrointest Surg. 2008;12: 1168–1176. 23. Yano F, El Sherif A, Filipi CJ, et al. Use of temporary esophageal stent in management of perforations after benign esophageal surgery. Surg Laparosc Endosc Percutan Tech. 2008;18:283–285. 24. Amrani L, M´enard C, Berdah S, et al. From iatrogenic digestive perforation to complete anastomotic disunion: endoscopic stenting as a new concept of “stent-guided regeneration and re-epithelialization.” Gastrointest Endosc. 2009;69:1282–1287. 25. Babor R, Talbot M, Tyndal A. Treatment of upper gastrointestinal leaks with a removable, covered, self-expanding metallic stent. Surg Laparosc Endosc Percutan Tech. 2009;19:1–4. 26. Freeman RK, Van Woerkom JM, Vyverberg A, et al. Esophageal stent placement for the treatment of spontaneous esophageal perforations. Ann Thorac Surg. 2009;88:194–198. 27. Salminen P, Gullichsen R, Laine S. Use of self-expandable metal stents for the treatment of esophageal perforations and anastomotic leaks. Surg Endosc. 2009;23:1526–1530. 28. Fernandez A, Vila JJ, Vazquez S, et al. Self-expanding plastic stents for the treatment of post-operative esophago-jejuno anastomosis leak. A case series study. Rev Esp Enferm Dig. 2010;102:704–710. 29. Kiernan PD, Khandhar SJ, Fortes DL, et al. Thoracic esophageal perforations. Am Surg. 2010;76:1355–1362. 30. Cerna M, Kocher M, Valek V, et al. Covered biodegradable stent: new therapeutic option for the management of esophageal perforation or anastomotic leak. Cardiovasc Intervent Radiol. 2011;34:1267– 1271. 31. Inbar R, Santo E, Subchi Ael-A, et al. Insertion of removable selfexpanding metal stents as a treatment for postoperative leaks and perforations of the esophagus and stomach. Isr Med Assoc J. 2011;13:230– 233. 32. van Boeckel PGA, Sijbring A, Vleggaar FP, et al. Systematic review: temporary stent placement for benign rupture or anastomotic leak of the oesophagus. Aliment Pharmacol Ther. 2011;33:1292–1301.
C 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
The Role of Esophageal Stents in the Management of Esophageal Anastomotic Leaks and Benign Esophageal Perforations Bobby V.M. Dasari, MRCS,∗ David Neely, MRCS,∗ Andrew Kennedy, FRCS,∗ Gary Spence, FRCS,† Paul Rice, FRCR,‡ Eamon Mackle, FRCS,‡ and Emmanuel Epanomeritakis, FRCS‡ Objective: The aim of this review was to assess the safety and effectiveness of esophageal stents in the management of benign esophageal perforation and in the management of esophageal anastomotic leaks. Background: Benign esophageal perforation and postoperative esophageal anastomotic leak are often encountered. Endoscopic placement of esophageal stent across the site of leakage might help control the sepsis and reduce the mortality and morbidity. Methods: All the published case series reporting the use of metallic and plastic stents in the management of postoperative anastomotic leaks, spontaneous esophageal perforations, and iatrogenic esophageal perforations were identified from MEDLINE, EMBASE, and PubMed (1990-2012). Primary outcomes assessed were technical success rates and complete healing rates. Secondary outcomes assessed were stent migration rates, stent perforation rates, duration of hospital stay, time to stent removal, and mortality rates. A pooled analysis was performed and subgroup analysis was performed for plastic versus metallic stents and anastomotic leaks versus perforations separately. Results: A total of 27 case series with 340 patients were included. Technical and clinical success rates of stenting were 91% and 81%, respectively. Stent migration rates were significantly higher with plastic stents than with metallic stents (40/148 vs 13/117 patients, respectively; P = 0.001). Patients with metallic stents had significantly higher incidence of postprocedure strictures (P = 0.006). However, patients with plastic stents needed significantly higher number of reinterventions (P = 0.005). Mean postprocedure hospital stay varied from 8 days to 51 days. There was no significant difference in the primary or secondary outcomes when stenting was performed for anastomotic leaks or perforations. Conclusions: Endoscopic management of esophageal anastomotic leaks and perforations with the use of esophageal stents is technically feasible. It seems to be safe and effective when performed along with mediastinal or pleural drainage. Esophageal stent can, therefore, be considered as a treatment option in the management of patients who present early after esophageal perforation or anastomotic leak with limited mediastinal or pleural contamination. Keywords: benign, esophagus, leaks, perforation, stents (Ann Surg 2014;259:852–860)
P
ostoperative anastomotic leak after esophageal surgery is encountered in 8% to 10% of patients undergoing esophageal resection.1,2 A Swedish national prospective study reported 8% anastomotic leak rates, with 20% associated mortality rate after intrathoracic resection and anastomosis for esophageal cancer.3 The From the ∗ Belfast Health and Social Care Trust, Belfast, United Kingdom; †Ulster Hospital, Belfast, United Kingdom; and ‡Craigavon Area Hospital, Craigavon, United Kingdom. Disclosure: The authors declare no conflicts of interest. Reprints: Bobby V. M. Dasari, MRCS, 15, The Blvd, Wellington Square, Belfast, BT7 3LW. E-mail: [email protected]. C 2014 by Lippincott Williams & Wilkins Copyright ISSN: 0003-4932/14/25905-0852 DOI: 10.1097/SLA.0000000000000564
852 | www.annalsofsurgery.com
England and Wales, esophagogastric cancer audit reported higher 30-day mortality rates in patients with anastomotic leak than those without leak (9% with leak vs 3.7% without leak; adjusted odds ratio: 2.82; 95% confidence interval: 1.59–5.01; P < 0.001).2 Early recognition and prompt management of leakage might help improve the outcome. Other causes of benign esophageal perforation include iatrogenic perforation after endoscopy, Boerhaave (spontaneous) perforation, and traumatic perforations due to foreign body or trauma. The principles of management of esophageal perforation and anastomotic leak include control of ongoing spillage from the esophagus, drainage of pleural and/or mediastinal cavities, intravenous broad-spectrum antibiotics, gastric decompression, and enteral or parenteral nutritional support. Sepsis is often controlled by surgical or percutaneous drainage of empyema, drainage of perianastomotic or mediastinal contamination, and with the use of intravenous antibiotics. The use of self-expanding stents in the palliation of symptomatic esophageal malignancies is well established and is effective.4 There is growing interest in the use of self-expanding stents in benign esophageal perforation. Endoscopic stenting is reported to be helpful to control sepsis and to allow early oral intake. However, placement of stent in a nonstrictured lumen could be associated with significant rates of stent migration, resulting in technical and clinical failure rates. In addition, endoscopic intervention for anastomotic leaks can be associated with further disruption of anastomosis, and other potential complications include perforation, bleeding, and strictures. We conducted a systematic review of the current literature to evaluate the success rates and safety of esophageal stents in the management of postoperative anastomotic leaks and benign esophageal perforations. Subgroup analysis was performed to assess the role of plastic and metallic stents and to assess the role of stents in patients with anastomotic leaks and perforations separately.
METHODS A systematic review of the literature was performed evaluating the role of esophageal stents in the management of postoperative anastomotic leaks and benign esophageal perforations.
Search Strategy and Study Selection A detailed electronic search was carried out from the following databases: MEDLINE, EMBASE, and PubMed. The search was performed using the Medical Subject Headings (MeSH): “esophagus,” “anastomotic leak,” OR “esophageal perforation,” “stents” combined with Boolean operator “AND.” Publications in English were included. All studies published from 1990 to 2012 were considered. Abstracts of potentially relevant publications based on the titles were reviewed, and case series reporting the outcomes of stent placement for postoperative anastomotic leaks and benign perforations were included. The reference list of all the included studies was hand-searched for further eligible studies that were then reviewed. Annals of Surgery r Volume 259, Number 5, May 2014
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 259, Number 5, May 2014
Authors of the case series that pooled the outcomes of patients with benign and malignant etiology were contacted by e-mail to request the data for patients with stenting for benign pathology separately. Where necessary, authors were contacted to clarify any additional information regarding follow-up duration, hospital stay, and postprocedure complications, if not explicit in the article. We identified 18 such articles and contacted the authors, of whom 4 responded with useful information that was included in the review. Study selection is presented as a flowchart in Figure 1.
Inclusion and Exclusion Criteria Prospective or retrospective case series with 3 or more than 3 patients in any single series evaluating the role of metallic or plastic esophageal stents in the management of esophagogastric anastomotic leaks, iatrogenic or traumatic esophageal perforations, and Boerhaave perforation were included.
Stents and Benign Esophageal Leak
Exclusion criteria applied were malignant perforations, anastomotic or benign strictures, stent for leakage from gastric staple line (sleeve gastrectomy), and gastrojejunal and colorectal anastomosis.
Data Extraction Two reviewers extracted all the data independently using paper data extraction pro forma. The information extracted from each study was year of publication, study design, inclusion criteria, exclusion criteria, number of participants in each study, type of stent used (metallic or plastic), number of stents used per patient, additional procedures performed, data on efficacy and safety of stents such as technical success rates, clinical success rates, failure rates, stent-related complication rates (stent migration, perforation, further dehiscence of anastomosis after stent insertion, bleeding), reintervention rates, mode of reintervention (endoscopic or surgical), and long-term complications such as anastomotic strictures and malignancy.
FIGURE 1. PRISMA flowchart illustrating search and data extraction history of articles on stenting of benign esophageal perforations and anastomotic leaks. C 2014 Lippincott Williams & Wilkins
www.annalsofsurgery.com | 853
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 259, Number 5, May 2014
Dasari et al
Primary outcomes evaluated were technical success rates and complete healing rates. Technical success was defined as being able to deploy 1 or more stents to occlude the site of leakage with no evidence of leakage of contrast at the end of the procedure. Clinical success was defined as complete healing of the perforation or leakage by placement of single or multiple stents irrespective of whether the stent was left in situ or was removed. The procedure was considered a clinical failure when the patient had to undergo surgical resection for persistent leak, died before confirming complete healing, or had a persistent fistula at the end of follow-up. Secondary outcomes were stent migration rates, reintervention rates, additional procedures performed, postprocedure perforation and bleeding, in-hospital mortality rates, time to stent removal, cost analysis, and long-term complications such as stricture or malignancy. Stent migration rates include the reported rates of stent dislodgement. Reintervention rates are those in which the patient needed repositioning of a stent (either because the stent had migrated or to control ongoing leakage), endoscopic placement of clips, or glue for partially healed leak and surgery for persistent leak. Removal of a dislocated stent by endoscopic or surgical means was not included in the reintervention rates. Additional procedures performed are defined as those performed to drain the sepsis or for providing nutritional support such as feeding jejunostomy or percutaneous endoscopic gastrostomy.
Subgroup analysis was performed to evaluate the role of metallic and plastic stents. Subgroup analysis was also performed in patient groups with anastomotic leaks and those with other causes of benign perforations. For subgroup analyses, studies reporting only the outcomes for the specific subgroups were included and those with mixed data (such as metallic and plastic stents) were not included.5,6
Statistical Analysis Pooled data analysis was performed to evaluate the primary and secondary outcomes. Where the data of certain outcomes are not provided, results are interpreted as a percentage of the number of patients from the included studies. Comparison of outcomes among subgroups was performed using 2 independent samples t test and test of equality of 2 proportions (SPSS and MINITAB).
RESULTS Twenty-seven studies were identified and included in the review. Study selection is presented as a flowchart illustrated in Figure 1. Details of the included studies and study characteristics are included in Tables 1 and 2. A total of 340 patients were included. Plastic stents were used in 12 studies that included 148 patients. Metallic stents were used in 12 studies that included 117 patients. Two studies with combined data using plastic and metallic stents in 70 patients5,6 and 1 study
TABLE 1. Details of the Included Studies Year
Type of Publication
Study Duration
Single Center
Total No. Patients
Cause of Esophageal Leak
Parajinen et Roy-Choudhury et al8 Siersema et al9 Evrard et al10 Gelbmann et al11 Hunerbein et al12 Koch et al13
1999 2001
Case series Case series
Not specified 1995–1999
Single center Single center
3 14
Iatro: 1, Boerhv: 1, Trauma: 1 Leaks: 14
2003 2004 2004 2004 2005
Case series Case series Case series Case series Case series
1999–2002 1999–2003 2002–2003 1998–2003 2001–2003
Single center Single center Single center Single center Single center
11 4 9 9 3
Langer et al14 Schubert et al15 Fischer et al16 Radecke et al17 Han et al18 Freeman et al19 Freeman et al20 Kim et al21
2005 2005 2006 2006 2006 2007 2007 2008
Case series Case series Case series Case series Case series Case series Case series Case series
2000–2003 2000–2003 1997–2004 2001–2005 2002–2005 Not specified Not specified 1998–2007
Single center Single center Single center Single center Multicenter Single center Multicenter Single center
24 12 15 9 8 17 21 17
Tuebergen et al22
2008
1999–2006
Single center
32
Iatro: 3, Leaks: 3, Boerhv: 5 Leaks: 4 Iatro: 3, Leaks: 5, Boerhv: 1 Leaks: 9 Iatro: 1, Leaks: 1, Postoperation fistula: 1 Leaks: 24 Leaks: 12 Spontaneous: 5, Iatro: 10 Iatro: 4, Leaks: 5 Leaks: 8 Iatro: 17 Leaks: 21 Iatro: 4, Leaks: 7, Fistulas: 2, Boerhv: 4 Leaks: 24, Iatro: 4, Boerhv: 4
Yano et al23 Amrani et al24 Babor et al25 Freeman et al26 Salminen et al27 Fernandez et al28 Kiernan et al29 van Heel et al6
2008 2009 2009 2009 2009 2010 2010 2010
Not specified 2005–2008 2004–2006 Not specified 1999–2008 2007–2008 1998–2009 2001–2009
Single center Single center Single center Single center Single center Single center Single center Single center
3 4 5 19 10 4 8 33
D’Cunha et al5 Cerna et al30 Inbar et al31
2011 2011 2011
2005–2009 2008–2010 2009–2010
Single center Single center Single center
37 5 4
Author al7
Nonrandomized observational study Case series Case series Case series Case series Case series Case series Case series Case series Case series Case series Case series
Iatro: 1, Leaks: 2 Perf: 2, Leaks: 2 Leaks: 4, Boerhv: 1 Spontaneous perfs: 19 Perfs: 8, Leak: 2 Leaks: 4 Perfs: 8 Iatrogenic: 19, Boerhv: 10, FB: 1, Trauma: 1, Leak: 1 Leaks: 22, Perfs: 15 Leaks: 4, Perfs: 1 Leaks: 3, Esophagopleural fistula: 1 (excluded patients with leak from sleeve gastrectomy)
Boerhv indicates Boerhaave; Iatro, iatrogenic perforations; Leaks, anastomotic leaks; Perfs, perforations.
854 | www.annalsofsurgery.com
C 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 259, Number 5, May 2014
Stents and Benign Esophageal Leak
TABLE 2. Characteristics of the Included Studies Metallic/ Plastic
Investigations to Confirm
Evrard et Fernandez et al28
P P
OGD, CS NM
Polyflex Polyflex
Freeman et al19
P
CS
Polyflex
Freeman et al20
P
CS
Polyflex
Freeman et al26
P
CS
Polyflex
Gelbmann et al11 Hunerbein et al12 Kim et al21
P P P
NM CS CS, CT
Langer et al14 Radecke et al17 Schubert et al15 Yano et al23 Amrani et al24 Babor et al25 Fischer et al16 Han et al18 Inbar et al31 Kiernan et al29 Koch et al13
P P P P M M M M M M M
NM NM OGD, CS CS NM NM OGD, CS, CT OGD, CS, CT OGD NM NM
Parajinen et al7 Roy-Choudhury et al8 Salminen et al27 Siersema et al9
M M
OGD, CS, CT CS
M M
OGD, CS, CT NM
Tuebergen et al22
Author al10
M
OGD
D’Cunha et al5 van Heel et al6
M, P M, P
OGD, CS, CT NM
Cerna et al30
BD
CS
Additional Procedures Performed
Type
Polyflex Polyflex Montgomery salivary bypass stents Polyflex Polyflex Polyflex Polyflex Niti-S: 3, Hanaro: 1 Covered Ella Boubella Ultraflex: 13, Niti-S: 2 Mushrom-shaped stents Sx-Ella, covered Covered wall stent Covered Dostent: 1, Hanaro: 1, Rosch: 1 Coated Schneider stent: 3 Ultraflex: 10, Gianturco: 3, Telestep: 4 Ultraflex: 1, Hanaro: 9 Wall stent: 7, Ultraflex: 4 Ultraflex Covered Ninitol/Polyflex Fer X Ella: 1, Flamingo: 3, Ultraflex: 35, Polyflex: 5, Hanarostent: 6 Ella–BD
Follow-up Period
NM Perianastomotic drainage: 2, tube thoracotomy: 2, radiological drain: 1 PEG: 12, esophageal dilatation: 1, thoracotomy: 4, VATS decortication: 2, celiotomy, gastrostomy, jejunostomy: 1 Tube jejunostomy: 10, thoracotomy: 5, PEG: 4, decortication: 4, celiotomy: 2, neck exploration: 1 PEG: 19, VAT decortication: 5, tube thoracotomy: 4 Chest drain in all: 9 Pleural drainage in all: 9 NM
NM Mean: 12 mo NM
NM Pleural/mediastinal drainage: 6 Pleural drainage: 14 Mediastinal drain: 1, gastrostomy: 1 None Surgical drainage: 5 Pleural drainage: 7, surgical: 2 Pleural drain: 8, jejunostomy: 3 None Chest drain: 1 Perianastomotic drainage: 1
Median: 220 d 40–565 d 9 mo 6–18 mo NM 4–12 mo 6 mo to 6 yr 7–30 mo NM Incomplete data 1–2 yr
Thoracotomy for drainage: 1 Pleural drainage: 3
Incomplete data 8 patient-months (range: 3–18 patient-months) NM Incomplete data
Pleural drainage: 4 Enteral feeding 2 patients, parenteral feeding 12 patients, pleural drainage 3 patients Pleural and perianastomotic drainage: 32 Drainage procedures: 19 Pleural drainage in all: 33
Gastrostomy: 1
8 mo NM >1 mo
>1 mo >3 mo
24 mo 41.6 mo Median: 2 yr
1–24 mo
BD indicates biodegradable; CS, contrast study; CT, contrast tomogram; M, metallic; NM, not mentioned; OGD, gastroscopy; P, plastic; PEG, percutaneous endoscopic gastrostomy; VATS, video-assisted thoracoscopic surgery.
reporting the use of biodegradable stents30 in 5 patients were included. Polyflex stents were the most often used plastic stents. Kim et al21 used Montgomery salivary bypass stents. Metallic stents were used in the remaining series. Subgroup analysis was performed to compare the role of plastic and metallic stents. In view of the limited numbers in the subgroups, we have not analyzed the covered and uncovered metallic stents. Nine case series with 100 patients reported use of stents for anastomotic leaks only; 4 case series with 54 patients reported use of stents for benign perforations only. Subgroup analysis was performed to compare the role of stents in these 2 groups. Diagnosis of perforation or anastomotic leak was made using contrast study, computed tomography, or gastroscopy. Contrast study was the commonly used mode of investigation whereas most studies used more than 1 modality of investigation. There was a wide variation C 2014 Lippincott Williams & Wilkins
in the duration of follow-up period among the included studies, from 1 month to 42 months (Table 2).
Primary Outcomes The primary outcomes evaluated were the technical and clinical success rates of esophageal stenting in benign perforation and anastomotic leakage.
Technical and Clinical Success Rates Technical success was achieved in 311 of 340 (91.4%) patients. Clinical success was achieved in 250 of 308 (81.1%) patients. Two studies14,29 did not report the complete healing rates, and these were excluded from the evaluation of clinical success rates. Patients with persistent leak or fistula (29 patients), those who died before confirmation of healing, and those lost to follow up (19 patients) from www.annalsofsurgery.com | 855
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 259, Number 5, May 2014
Dasari et al
the available data were considered as clinical failures to heal. Reason for clinical failure is not clear in 10 patients. Technical success rates in the plastic and metallic stent groups were 89.8% versus 96.5% (P = 0.025). Clinical success rates were similar in both the groups (86.2%). A total of 404 stents were used in 300 patients (1.34 stents per patient). A total of 174 plastic stents were used for 148 patients (1.17 plastic stents per patient), 83 metallic stents were used for 77 patients (1.07 metallic stents per patient), and 7 biodegradable stents were used for 5 patients (1.4 per patient). Two studies5,6 provided combined data of plastic and metallic stents. It was unclear from 2 studies about how many stents were used (40 patients).22,29
Secondary Outcomes
to healing is assessed after removal of stent, and it has, therefore, varied significantly between the studies.
Mortality Rates Forty-four deaths were reported, and it was not clear from the included studies whether the mortality was in-hospital, 30-day, or 90-day mortality rates. Also, not all of the deaths were as a result of perforation or sepsis related to perforation (Table 3).
Long-term Complications Eleven postprocedure strictures were reported in 340 patients (1 in plastic stents vs 9 in metallic stents; P = 0.006). No malignancies were reported in any of the series at follow-up.
Stent Migration
Leaks Versus Perforations
The reported overall stent migration rates were 73 of 340 patients (20.8%), 40 of 148 patients (27%) with plastic stents, and 13 of 117 patients (11%) with metallic stents (P = 0.001) (Table 3).
This subgroup analysis was performed from the studies that have reported these results separately (Table 4). There were no significant differences in the primary or secondary outcomes evaluated.
Stent Perforation Stent perforation was reported in 7 of 340 patients (2%) (plastic vs metallic: 2 vs 5; P = 0.69) (Table 3). Two patients were reported to have suffered from bleeding because of erosion of the stent into the aorta. Further dehiscence of the anastomosis was reported in 1 patient.
Stent-Related Bleeding Three patients were reported to have suffered bleeding after stent deployment.
Reintervention Rates Endoscopic reinterventions were reported in a total of 58 patients (17%) and surgical reintervention was reported in 33 patients (9.7%). Where the data were provided separately for the plastic and metallic stents, overall reintervention rates were significantly higher in the plastic stent group than in the metallic stent group (41 vs 13; P = 0.005). Endoscopic reinterventions were required in 22% versus 5% and surgical intervention was required in 5.8% versus 15% between the plastic and metallic groups (P < 0.00001 and 0.14, respectively).
Additional Procedures Performed The role of esophageal stenting in the management of patients with esophageal perforation or anastomotic leaks is to limit the sepsis from continued leak and to allow early resumption of enteral feeding to attend the nutritional needs. However, most of these patients need additional procedures for the management of established or coexisting sepsis in the pleural cavity or mediastinum. Thoracotomy (tube, surgical or video-assisted) was the most common additional procedure performed, reported in 187 patients. The use of percutaneous gastrostomy (35 patients), surgical gastrostomy (2 patients), jejunostomy (13 patients), and parenteral nutrition (12 patients) for nutritional support has been reported (Table 2).
Duration of Hospital Stay Duration of hospital stay was reported in 13 of the included studies. It varied from mean postoperative hospital stay of 8 days to 51 days, reflecting the complex and prolonged course of recovery.
Time to Stent Removal and Time to Healing There is significant variability in the reported studies regarding the time to stent removal (Table 3). Although some centers removed the stents at 4 weeks, others have left the stent for longer period. Time 856 | www.annalsofsurgery.com
DISCUSSION Twenty-seven case series reported the use of stents in the management of esophageal anastomotic leaks or benign esophageal perforations, and these studies represent the current available evidence on this subject. Currently, there are no published studies comparing the use of stents versus surgical or conservative management, nor comparing the role of metallic versus plastic stents for this group of patients. van Boeckel et al32 in their systematic review included patients with anastomotic dehiscence after bariatric surgery in addition to the patient groups included in our review and the data published until 2010. Patients who underwent bariatric surgery are not included in our review. All the remaining articles (an overlap of 18 studies between the reviews), other relevant studies (3), and the larger series published up to 2012 (6 additional studies) are included. Appropriate preprocedure workup of esophageal perforation typically includes a contrast study. Computed tomography allows evaluation of the severity of mediastinal or pleural space contamination. Endoscopic visualization helps identify the exact location of the perforation, extent of dehiscence, and the viability of esophageal mucosa. Investigations used in the preoperative workup among the included studies are listed in Table 2. Contrast study is the commonly used mode of investigation whereas most studies used more than 1 modality of investigation. The overall technical and clinical success rates of esophageal stent placement in these patient groups were 91% and 81%, respectively, similar to previously published data.32 The procedure is technically feasible, with most individual case series reporting 100% technical success rates. It is important to know whether the technical success is being translated to clinical success (complete healing) and the overall desired outcomes of the procedure (such as time to complete healing and duration of hospital stay). The variables that determine overall success might include complete healing of the leak, control of sepsis after stent deployment, lack of or limited reintervention rates, early return to oral intake, and low overall morbidity and mortality rates. Complete healing rates of the leak are certainly impressive (81%), and these are comparable with previously published reviews. There is no statistically significant difference in the complete healing rates between the plastic and metallic stents (P = 0.99). The most common source of sepsis in this patient group is pleural, mediastinal, or perianastomotic contamination, and 63% of patients required a form of drainage procedure for controlling sepsis. Severity of sepsis after esophageal perforation might also be determined by the delay associated with identification of leak. Fischer C 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
C 2014 Lippincott Williams & Wilkins
P M
M M
M
Yano et al23 Amrani et al24
Babor et al25 Fischer et al16
Han et al18
4 8
8
5 15
3 4
12
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Salminen et
M
10
P
Schubert et al15
9
14
P
Radecke et al17
24
al27
P
Langer et al14
9 17
Roy-Choudhury et al8 M
P P
Hunerbein et al12 Kim et al21
9
3 3
P
Gelbmann et al11
19
M M
P
Freeman et al26
21
Koch et al13 Parajinen et al7
P
Freeman et al20
17
M M
P
Freeman et al19
4 4
Inbar et al31 Kiernan et al29
P P
Fernandez et al28
Evrard et al10
Author
5
1
0 0
0 2
1
0 2
0 0
0
4
6
0 1
3
0
1
0
0 1
0
0 0
0 0
0
0
2
0 0
0
0
0
0
0 0
1
0 0
0
2 0 1 1 (eroded into thoracic aorta) 3 0 0 0
0
2 0
1 0
2
2
9
2 6
3
4
5
3
3 0
0
1
0 0
0 0
0
0 0
0 0
0
0
0
0 0
0
0
0
0
0 0
Restent: 2
Restent: 2, repositioning: 4
Restent/ repositioning: 4 Restent: 1, repositioning: 3 Restent: 2 Repositioning: 4
Restent/ repositioning: 3 Restent/ repositioning: 5
2
0
2 0
1 4
0
0 0
1 1
0 0
Repositioned: 1
0
1 0
Repositioned: 1 Unclear
0
0 Restent + clips + glue: 1 0 0
1
0
1 0
0 1
0
0 0
1 0
0
0
3
0 0
0
2
1
1
0 0
1
0
1 1
0 0
1
0 2
0 0
0
0
0
0 0
0
0
1
0
0 0
0
0
0 0
0 0
0
0 0
0 0
0
0
0
0 0
0
0
0
0
0 0
13
14
4 3
4 Unclear
8
5 15
3 5
12
11
27
18 17
10
21
27
19
4 5
8
14
3 3
4 7
8
5 15
3 4
12
9
16
8 17
7
17
20
16
4 4
Reason for Clinical Failure
Time to Stent Removal
... Not mentioned 2 mortality, 1 lost 4 wk to follow up—so clinical success could not be confirmed 8 ... 18–48 d; mean 32 ± 11 d 4 ... 6–7 wk Unclear Outcome after 4–6 wk stent removal is not mentioned 3 ... Unclear 3 ... Not removed the stents 13 Persistent leak and Unclear death: 1 8 Persistent leak and 3–16 wk; death: 1, persistent median: 10 wk leak requiring surgical resection: 1 (continued)
5 12
... 1 wk to 4 mo Early mortality: 1, 4–8 wk minor persistent fistula: 1 16 Continued leak req Range: 17–91 d; surgery: 1 52 ± 20 d 20 Complete Range: 15–175 d; dehiscence of mean: 51 ± 43 d existing surgical repair due to stent placement: 1 17 Persistent leak: 2 Range: 14–32 d; mean: 20 ± 15 d 6 Death: 3 Range: 32–242 d; mean: 135 ± 78 d 9 ... mean: 29 d 13 Persistent fistula: 3 Range: 12-109d; mean: 38 ± 25 d Unclear 4 were not assessed Unclear by postcontrast study 6 Death: 3 Range: 15–438 d; median: 55.5 d 11 Persistent leak: 1 Range: 2–8 wk; median: 4 wk 3 ... 5–10 wk 4 ... Average: 5 wk
4 2
Surgery as Total Restent/ Reinterven- Strict- Malign- No. Technical Clinical Repositioning tion ure ancy Stents Success Success
UnclearUnclear
2
9
2 4
4
6
8
5
Unclear 0
Total Number Metallic/ No. of Migra- Number of ReinterPlastic Patients Mortality tions Perforations Bleeding ventions
TABLE 3. Primary and Secondary Outcomes of Patients Managed With Plastic (P) Stents and Metallic (M) Stents∗
Annals of Surgery r Volume 259, Number 5, May 2014 Stents and Benign Esophageal Leak
www.annalsofsurgery.com | 857
Annals of Surgery r Volume 259, Number 5, May 2014
5 7 0 0 0 1 1 0 0 3 0 5 BD
Two studies reported combined data with both (B) plastic and metallic stents and 1 with biodegradable (BD) stents. BD indicates biodegradable; M, metallic; P, plastic.
Cerna et al30
∗
11 B van Heel et al6
858 | www.annalsofsurgery.com
33
37 B D’Cunha et al5
7
4
4 required surgery 5 wk (R: 1–84 wk) for stent failure, 7 deaths (total 10) Persistent leak: 1 ... 23 32 46 0 1 4 Repositioning: 4, restent: 4 12 2
22 28 94 0 0 9 16 25 0
3 (gastroaortic fistula: 1) 1 6
0 0 3 5 32 Tuebergen et al22 M
5
8
3
0
Unclear
32
25
Technical failure 1, Median: 7 wk sepsis due to (R: 6–14 wk) abscess 1: both requiring esophagectomy Unclear Median: 46 d (R: 4–426 d) Nonseal: 9 patients Not mentioned 9 10 12 0 0 2
Endoscopy and removal of metallic strand causing dysphagia: 1 UnclearUnclear 3 0 0 1 0 11 M Siersema et al9
Author
Total Number Metallic/ No. of Migra- Number of ReinterPlastic Patients Mortality tions Perforations Bleeding ventions
TABLE 3. (Continued)
Surgery as Restent/ Reinterven- StrictRepositioning tion ure
Total Malign- No. Technical Clinical ancy Stents Success Success
Reason for Clinical Failure
Time to Stent Removal
Dasari et al
et al16 reported that stent insertion within 1 hour resulted in uneventful recovery, with average postoperative hospital stay of 5 days compared with patients who had a stent inserted with a median delay of 123 hours. In the delayed group, the clinical course was complicated with multiorgan failure, and the average postoperative hospital stay was 44 days.16 None of the included studies reported the severity of sepsis at the time of endoscopic intervention. Successful deployment of stent controlling the leakage or spillage should ultimately result in complete healing provided the stent remains in the required position until the dehiscence heals. Yakoub et al4 reported migration rates of 5% versus 10% (P = 0.09) in their meta-analysis comparing the use of metallic and plastic stents for patients with malignant dysphagia. With no stricture and hinge to keep the stent in place, the migration rates are expected to be higher with its use in benign nonstricturing etiology. Overall stent migration rates in our review are 20.8% and are significantly better with the metallic stents (P = 0.001). Despite the reported high migration rates, clinical healing rates are, however, similar between the 2 groups. This could be either due to migration of the stent after complete healing or due to reintervention undertaken after migration of the stent. The most common endoscopic reintervention is placement of another stent due to persistent leak, stent migration, or repositioning of the stent. Endoscopic placement of fistula plugs or resolution clips for persistent fistula is also reported. From this review, 13% mortality rates with the endoscopic management of perforations and leaks seem to be more favorable than the mortality rates reported with surgical intervention. However, we are not able to ascertain whether these are 30-day, 90-day, or even long-term mortality rates from the reported studies. From the available studies, it is difficult to evaluate the time taken for the perforation or leak to heal. Time to heal can be accurately ascertained only after removing the stent, followed up by a contrast study or direct visualization. Difficulty to remove stents after resolution of the leak/perforation due to hyperplasia of the tissue in the uncovered parts at both ends is reported and can induce stenosis that is a concern. Animal studies have suggested that 4 weeks should be sufficient time for tissue healing. Although the more recent studies have opted to remove the stent after 6 to 8 weeks,5,15,23,24,27,28,31 some of the earlier studies left the stent in situ much longer or even permanently. Another reason for ambiguity regarding the time to healing is whether the stent removal has caused disruption of healed mucosa, resulting in the leak of contrast during check contrast study. Cost considerations are also important when evaluating the 2 different intervention strategies. None of the included case series performed cost-effectiveness analysis. In an attempt to reduce the heterogeneity, we have included only those patients with anastomotic leaks and benign perforations. However, heterogeneity still remains a drawback of this current review, as it includes varied patient groups such as those with iatrogenic perforations after dilatation of a benign stricture and anastomotic leak after cancer resection. Outcomes of the patients with significant mediastinal or pleural contamination are expected to be worse than those with minimal contamination and lack of this differentiation adds to the heterogeneity of the patient groups included in the study. Although this review is based on case series with small patient numbers, it provides the best available evidence on this topic and suggests that endoscopic stent placement is a safe, acceptable and effective minimally invasive procedure to manage patients with benign esophageal leaks. The review, however, does not provide comparative data between the standard surgical management and the novel endoscopic management and covered or uncovered stents. The role of stent in patients with significant mediastinal or pleural contamination and the time to removal of stent remain unclear. There is no evidence to suggest that stent placement is superior to other forms of standard C 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
C 2014 Lippincott Williams & Wilkins
P
P P
P
Freeman et al26
Fischer et al16 Freeman et al19
Parajinen et al7
L indicates leak; P, perforation.
L
Inbar et al31
3 54
15 17
4 100 19
4 4 8 9 12 24
21
L
L L L L L L
14
L
0 2
2 0
0 10 0
0 1 1 0 0 6
1
1
0 7
0 3
2 23 4
3 0 0 2 2 9
5
0
0 0
0 0
0 2 0
0 0 0 0 0 2
0
0
0 11
0 5
1 20 6
Unclear 0 0 2 Unclear 9
8
0
Leak/ Total No. Mortality Stent Migration Stent Perforation Reintervention Perforation Patients Rates Rates Rates Rates
Evrard et Fernandez et al28 Han et al18 Hunerbein et al12 Schubert et al15 Langer et al14
al10
Roy-Choudhury et al8 Freeman et al20
Author
Restent/ repositioning: 4 0 Restent/ repositioning: 3 0
Restent/ repositioning: 5 0 0 0 Restent: 2 Unclear Restent: 2, repositioning: 4 Repositioned: 1
0
Restent
0
0 1
2
0
0 0 0 0 0 3
1
0
1 3
2 0
0 2 0
0 0 1 0 0 0
1
0
3 58
15 19
4 119 21
4 5 8 18 12 27
27
14
3 51
15 16
4 90 17
4 4 8 8 12 16
20
14
3 48
12 16
4 71 17
4 2 8 9 11 Unclear
20
13
Surgery as Total No. Technical Clinical Reintervention Stricture Stents Success Success
TABLE 4. Subgroup Analysis Comparing the Outcomes in the Anastomotic Leak Group Versus Perforations Group
Annals of Surgery r Volume 259, Number 5, May 2014 Stents and Benign Esophageal Leak
www.annalsofsurgery.com | 859
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 259, Number 5, May 2014
Dasari et al
treatment due to the lack of randomized controlled trials. There is a role to perform randomized controlled trials in homogeneous patient groups comparing the surgical and endoscopic management and various types of stents with well-defined outcomes including clinical success rates and to assess the optimum time for stent removal.
CONCLUSIONS Endoscopic management of esophageal anastomotic leaks and benign perforations is technically feasible. It seems to be a safe and effective management strategy when performed along with an additional drainage procedure to control sepsis. Esophageal stent can, therefore, be considered as a treatment option in the management of patients who present early after esophageal perforation or anastomotic leak with limited mediastinal or pleural contamination.
REFERENCES 1. Viklund P, Linblad M, Lu M, et al. Risk factors for complications after esophageal cancer resection. A prospective population-based study in Sweden. Ann Surg. 2006;243:204–211. 2. http://www.augis.org/pdf/NHS-IC-OGC-Audit-2010-interactive.pdf. Accessed January 31, 2013. 3. Rutegard M, Lagergren P, Rouvelas I, et al. Intrathoracic anastomotic leakage and mortality after esophageal cancer resection: a population-based study. Ann Surg Oncol. 2012;19:99–103. 4. Yakoub D, Fahmy R, Athanasiou T, et al. Evidence-based choice of esophageal stent for the palliative management of malignant dysphagia. World J Surg. 2008;32:1996–2009. 5. D’Cunha J, Rueth NM, Groth SS, et al. Esophageal stents for anastomotic leaks and perforations. J Thorac Cardiovasc Surg. 2011;142:39–46. 6. van Heel NC, Haringsma J, Spaander MC, et al. Short-term esophageal stenting in the management of benign perforations. Am J Gastroenterol. 2010;105:1515–1520. 7. Parajinen J, Ristkari SK, Mokka REM. A report of three cases with an oesophageal perforation treated with a coated self-expanding stent. Ann Chir Gynaecol. 1999;88:332–334. 8. Roy-Choudhury SH, Nicholson AA, Wedgwood KR, et al. Symptomatic malignant gastroesophageal anastomotic leak: management with covered metallic esophageal stents. Am J Roentgenol. 2001;176:161–165. 9. Siersema PD, Homs MYV, Haringsma J, et al. Use of large diameter metallic stents to seal traumatic non-malignant perforations of the esophagus. Gastrointest Endosc. 2003;58:356–361. 10. Evrard S, Le Moine O, Lazaraki G, et al. Self-expanding plastic stents for benign esophageal lesions. Gastrointest Endosc. 2004;60:894–900. 11. Gelbmann CM, Ratiu NL, Rath HC, et al. Use of self-expandable plastic stents for the treatment of esophageal perforations and symptomatic anastomotic leaks. Endoscopy. 2004;36:695–699. 12. H¨unerbein M, Stroszczynski C, Moesta KT, et al. Treatment of thoracic anastomotic leaks after esophagectomy with self-expanding plastic stents. Ann Surg. 2004;240:801–807. 13. Koch S, Weber A, Fein F, et al. Esophageal stents as a salvage therapy for non-malignant iatrogenic esophageal perforations. Gastroenterol Clin Biol. 2005;29:735–739. 14. Langer FB, Wenzl E, Prager G, et al. Management of postoperative esophageal leaks with the Polyflex self-expanding covered plastic stent. Ann Thorac Surg. 2005;79:398–403.
860 | www.annalsofsurgery.com
15. Schubert D, Scheidbach H, Kuhn R, et al. Endoscopic treatment of thoracic esophageal anastomotic leaks by using silicone-covered, self-expanding polyester stents. Gastrointest Endosc. 2005;61:891–896. 16. Fischer A, Thomusch O, Benz S, et al. Nonoperative treatment of 15 benign esophageal perforations with self-expandable covered metal stents. Ann Thorac Surg. 2006;81:467–472. 17. Radecke K, Lang H, Frilling A, et al. Successful sealing of benign esophageal leaks after temporary placement of a self-expanding plastic stent without fluoroscopic guidance. Z Gastroenterol. 2006;44:1031– 1038. 18. Han XW, Li YD, Wu G, et al. New covered mushroom-shaped metallic stent for managing anastomotic leak after esophagogastrostomy with a wide gastric tube. Ann Thorac Surg. 2006;82:702–706. 19. Freeman RK, Van Woerkom JM, Ascioti AJ. Esophageal stent placement for the treatment of iatrogenic intrathoracic esophageal perforation. Ann Thorac Surg. 2007;83:2003–2007. 20. Freeman RK, Ascioti AJ, Wozniak TC. Postoperative esophageal leak management with the Polyflex esophageal stent. J Thorac Cardiovasc Surg. 2007; 133:333–338. 21. Kim AW, Liptay MJ, Snow N, et al. Utility of silicone esophageal bypass stents in the management of delayed complex esophageal disruptions. Ann Thorac Surg. 2008;85:1962–1967. 22. Tuebergen D, Rijcken E, Mennigen R, et al. Treatment of thoracic esophageal anastomotic leaks and esophageal perforations with endoluminal stents: efficacy and current limitations. J Gastrointest Surg. 2008;12: 1168–1176. 23. Yano F, El Sherif A, Filipi CJ, et al. Use of temporary esophageal stent in management of perforations after benign esophageal surgery. Surg Laparosc Endosc Percutan Tech. 2008;18:283–285. 24. Amrani L, M´enard C, Berdah S, et al. From iatrogenic digestive perforation to complete anastomotic disunion: endoscopic stenting as a new concept of “stent-guided regeneration and re-epithelialization.” Gastrointest Endosc. 2009;69:1282–1287. 25. Babor R, Talbot M, Tyndal A. Treatment of upper gastrointestinal leaks with a removable, covered, self-expanding metallic stent. Surg Laparosc Endosc Percutan Tech. 2009;19:1–4. 26. Freeman RK, Van Woerkom JM, Vyverberg A, et al. Esophageal stent placement for the treatment of spontaneous esophageal perforations. Ann Thorac Surg. 2009;88:194–198. 27. Salminen P, Gullichsen R, Laine S. Use of self-expandable metal stents for the treatment of esophageal perforations and anastomotic leaks. Surg Endosc. 2009;23:1526–1530. 28. Fernandez A, Vila JJ, Vazquez S, et al. Self-expanding plastic stents for the treatment of post-operative esophago-jejuno anastomosis leak. A case series study. Rev Esp Enferm Dig. 2010;102:704–710. 29. Kiernan PD, Khandhar SJ, Fortes DL, et al. Thoracic esophageal perforations. Am Surg. 2010;76:1355–1362. 30. Cerna M, Kocher M, Valek V, et al. Covered biodegradable stent: new therapeutic option for the management of esophageal perforation or anastomotic leak. Cardiovasc Intervent Radiol. 2011;34:1267– 1271. 31. Inbar R, Santo E, Subchi Ael-A, et al. Insertion of removable selfexpanding metal stents as a treatment for postoperative leaks and perforations of the esophagus and stomach. Isr Med Assoc J. 2011;13:230– 233. 32. van Boeckel PGA, Sijbring A, Vleggaar FP, et al. Systematic review: temporary stent placement for benign rupture or anastomotic leak of the oesophagus. Aliment Pharmacol Ther. 2011;33:1292–1301.
C 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
![[The role of stents in the treatment of anastomotic leaks and esophageal perforation].](/assets/img/hold.png)
