546435 research-article2014
SJS104010.1177/1457496914546435Treatment of esophageal perforationA. Troja, et al.
Original Article
Scandinavian Journal of Surgery 104: 191–195, 2014
Treatment Of Esophageal Perforation: A Single-Center Expertise A. Troja, P. Käse, N. El-Sourani, S. Miftode, H. R. Raab, D. Antolovic Department for General and Visceral Surgery, Klinikum Oldenburg, Oldenburg, Germany
Abstract
Background and Aims: Esophageal perforation is a rare diagnosis, which is associated with a high morbidity and mortality. There is only small scientific background regarding the best choice of treatment. Parameters indicating a good clinical outcome seem to be localization, depth of the defect, pre-existing risk factors, and time interval between the event and start of treatment. Material and Methods: We evaluate retrospective data from 39 patients who were treated with a esophageal perforation in our hospital between 2004 and 2012. Results and Conclusions: Our collected data agree with the available published literature. Endoscopic treatment seems to be favorable in early diagnosis. Key words: Esophageal perforation; stenting; esophagectomy
Introduction Esophageal perforation is a potential life-threatening diagnosis of various etiologies (1, 2). Due to a low number of cases describing such an event, an evidence-based guideline for treating esophageal perforation is not available. De Schipper et al. (3) tried to create a treatment algorithm for Boerhaave’s Syndrome by means of a retrospective analysis by collecting data from 1974 to 2008. In 2011, Soreide et al. established a therapeutic and diagnostic algorithm for the initial treatment in the trauma room and for the following 24 h. (4) Brinster et al. (5) were able to show that in 59% of the cases, the etiology is iatrogenic, in most cases due to upper endoscopy. The total risk of a perforation caused by a diagnostic flexible upper endoscopy is
Correspondence: Achim Troja Department for General and Visceral Surgery Klinikum Oldenburg Rahel-Strauss-Str. 10 26133 Oldenburg Germany Email: [email protected]
described as low as 0.03% but increases with further interventions (6). Esophageal perforation tends to occur at four predisposing anatomical positions: (a) the Killian Triangle, a muscle-free zone of the proximal esophagus; (b) at the crossing at the aortic arch; (c) at the connection to the left main bronchus; and (d) at the gastroesophageal junction. Regarding the etiology, esophageal perforation can also be caused due to a malignancy or in the setting of the above-mentioned Boerhaave’s Syndrome. The latter describes a barotrauma due to regurgitation, which leads to left lateral perforation of the esophagus (3). Due to the anatomical conditions in the mediastinum with absence of immunocompetent tissue, the body is unable to cover the defect, leading to consecutive mediastinitis with a mortality up to 20% (1). Thus, a prompt and targeted diagnosis and therapy are crucial. The localization, extent of the defect, and the time gap between the event and the initiation of therapy can be used in the decision-making process, although no evidence-based recommendations are available in the published literature. In this analysis, we updated the formerly published data (7) by complementing it with further collective data from 2008 to 2012. The aim of our study was to
192
A. Troja, et al.
show that the early diagnosis seems to be the main impact factor for successful treatment. In fact, an early diagnosis also favors an interventional endoscopical treatment. Methods And Material Between 2004 and 2012, a total of 39 patients (20 women and 19 men) with the diagnosis of an esophageal perforation have been treated at the Klinikum Oldenburg. The mean age at diagnosis was 66.4 years (range: 16–91 years). The data were retrospectively evaluated (Table 1). Results Etiology, Localization, And Diagnostics
Regarding the etiology of esophageal perforation, four groups exist: (a) 18/39 iatrogenic/interventional-associated (2 transesophageal echocardiography (TEE), 1 cystic duct stent replacement, 1 endoscopic retrograde cholangio pancreatography (ERCP), 1 polypectomy of the duodenum, 1 submucosal dissection of Barrett’s esophagus, 6 dilatations, 3 diagnostic endoscopies), (b) 9/39 accidental/foreign body ingestion, (c) 4/39 Boerhaave’s Syndrome, and (d) 3/39 postoperative and 5 events could not be classified. A safe inquiry about the presenting symptoms could not be obtained from the available data. We refer to our initial analysis in which one-quarter of the patients presented with leading symptoms suggestive of esophageal perforation (7). The most common localization of an esophageal perforation was the distal third with 16/29, followed by the middle third 14/39, and the proximal third with 9/39 patients. An upper endoscopy and a computed tomography (CT) scan with contrast were performed in all patients. The mean size of the defect was 2.4 cm (range: 5 cm, n = 19; however, size of the defect was not documented in all patients). Treatment
Of the 39 patients, 17 were referred to us from the surrounding hospital which partially performed frustrating therapeutic attempts, 9 patients presented to our surgical emergency room, and 13 patients had already been treated internally. Treatment was initiated within 12 h of the event in 18/39 patients, at >12 h in 5/39 patients, and at >24 h in 16/39 patients (Table 2). Conservative/interventional treatment A total of 22/39 patients were treated conservatively or interventional. Of the 22 patients, 7 were treated by stenting, 1 patient was treated by application of hemoclips, and 14 were treated by observation and/ or recurrent endoscopic lavages. In five patients, an interventional procedure was followed by surgery. Out of those patients, two received surgery, as we were unable to resect (R0) the adenocarcinoma endoscopically. Thus, resection of the malignancy was
performed after primary stabilization with a stent. In n = 1, the state of the patient deteriorated after initial stenting, leading to surgery. In n = 2, initial stenting with subsequent surgery was the treatment plan. Insertion of a thoracic drainage was not performed routinely in the interventional group, rather dependent on the CT result. Surgical treatment In n = 17 patients, an initial surgical procedure was performed. Different surgical techniques were used. In n = 5, stitching of the defect was performed; in n = 3, stitching of the defect with subsequent fundoplicatio was performed, of which n = 1 received interposition of the colon in a second setting, n = 1 received esophagectomy with gastric interposition in a second setting, n = 1 received esophagectomy with gastric interposition; n = 3 received esophagectomy with colon interposition; n = 2 received discontinuity resection of the proximal stomach and distal esophagus without reconstruction; in n = 1, abdominal approach through a right laparotomy, transhiatal lavage, and drainage were performed; and in n = 1, a collar approach with mediastinotomy and drainage was performed. The latter approach was used in treating a defect of the proximal esophagus after the extraction of glass fragments. In n = 3 patients who received an interposition of the colon, a gastric interposition was not performed because in n = 1, acid ingestions injured the stomach and in n = 1, gastric/esophageal gangrene occurred post fundoplicatio. Outcome
A total of n = 31 survived after esophageal perforation and n = 8 passed away, correlating to a mortality of about 21%. While looking at the patient collective according to the time passed before treatment was initiated, one notices that the survival rate was at 94% within the early treatment group (12 h). This outcome correlates with available published literature. Out of n = 8 who passed away, n = 1 was unable to receive surgical treatment due to her reduced state of health. Another patient was referred to us with a fulminant sepsis who deceased shortly after. Another patient passed away due to an endocarditis and a pulmonary embolism secondary to Boerhaave’s Syndrome. N = 2 deceased postoperatively due to a reduced state of health although no clear focus of infection could be identified. One died spontaneously due to an underlying hematothorax secondary to a pulmonary artery bleeding. One patient passed away due to malignant cardiac arrhythmia. Discussion This retrospective analysis was done to characterize and describe patients with an esophageal perforation, who in part were referred, diagnosed, and received their initial treatment from surrounding hospitals. Due to its etiology and acute clinical presentation, current data from random-control studies do not exist.
66 79 16 48 50 75 74 80 86 86 64
80 65 62 60 68 56 91
64 82 85
72 69 89 64
77 64
81 60
85 54 44 49 58 65
61
53 47
59
1 2 3 4 5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21
22 23 24 25
26 27
28 29
30 31 32 33 34 35
36
37 38
39
m
w w
w
w w m m m m
w m
w w
m w w m
w w w
w m m w m m w
m m m m m w w w m w m
Sex
Middle
Distal Middle
distal
Proximal cardia Middle Proximal Distal Proximal
Middle Middle
Middle Middle
Proximal Distal Middle Distal
Cardia Cardia middle
Proximal Cardia Middle Proximal Distal Distal Proximal
proximal Middle Middle Distal Distal Distal Middle Proximal Distal Middle Complete
Defect localisation
0,5 cm 1)0,2 cm 2)0,3 cm 0,6 cm
n.k.
n.k. 1)1,5cm, 2) 2cm n.k. 1,5 cm o.a. 0,8 cm n.k. n.k.
1 cm n.k.
n.k. n.k. n.k. n.k.
n.k. n.k.
SJS104010.1177/1457496914546435Treatment of esophageal perforationA. Troja, et al.
Original Article
Scandinavian Journal of Surgery 104: 191–195, 2014
Treatment Of Esophageal Perforation: A Single-Center Expertise A. Troja, P. Käse, N. El-Sourani, S. Miftode, H. R. Raab, D. Antolovic Department for General and Visceral Surgery, Klinikum Oldenburg, Oldenburg, Germany
Abstract
Background and Aims: Esophageal perforation is a rare diagnosis, which is associated with a high morbidity and mortality. There is only small scientific background regarding the best choice of treatment. Parameters indicating a good clinical outcome seem to be localization, depth of the defect, pre-existing risk factors, and time interval between the event and start of treatment. Material and Methods: We evaluate retrospective data from 39 patients who were treated with a esophageal perforation in our hospital between 2004 and 2012. Results and Conclusions: Our collected data agree with the available published literature. Endoscopic treatment seems to be favorable in early diagnosis. Key words: Esophageal perforation; stenting; esophagectomy
Introduction Esophageal perforation is a potential life-threatening diagnosis of various etiologies (1, 2). Due to a low number of cases describing such an event, an evidence-based guideline for treating esophageal perforation is not available. De Schipper et al. (3) tried to create a treatment algorithm for Boerhaave’s Syndrome by means of a retrospective analysis by collecting data from 1974 to 2008. In 2011, Soreide et al. established a therapeutic and diagnostic algorithm for the initial treatment in the trauma room and for the following 24 h. (4) Brinster et al. (5) were able to show that in 59% of the cases, the etiology is iatrogenic, in most cases due to upper endoscopy. The total risk of a perforation caused by a diagnostic flexible upper endoscopy is
Correspondence: Achim Troja Department for General and Visceral Surgery Klinikum Oldenburg Rahel-Strauss-Str. 10 26133 Oldenburg Germany Email: [email protected]
described as low as 0.03% but increases with further interventions (6). Esophageal perforation tends to occur at four predisposing anatomical positions: (a) the Killian Triangle, a muscle-free zone of the proximal esophagus; (b) at the crossing at the aortic arch; (c) at the connection to the left main bronchus; and (d) at the gastroesophageal junction. Regarding the etiology, esophageal perforation can also be caused due to a malignancy or in the setting of the above-mentioned Boerhaave’s Syndrome. The latter describes a barotrauma due to regurgitation, which leads to left lateral perforation of the esophagus (3). Due to the anatomical conditions in the mediastinum with absence of immunocompetent tissue, the body is unable to cover the defect, leading to consecutive mediastinitis with a mortality up to 20% (1). Thus, a prompt and targeted diagnosis and therapy are crucial. The localization, extent of the defect, and the time gap between the event and the initiation of therapy can be used in the decision-making process, although no evidence-based recommendations are available in the published literature. In this analysis, we updated the formerly published data (7) by complementing it with further collective data from 2008 to 2012. The aim of our study was to
192
A. Troja, et al.
show that the early diagnosis seems to be the main impact factor for successful treatment. In fact, an early diagnosis also favors an interventional endoscopical treatment. Methods And Material Between 2004 and 2012, a total of 39 patients (20 women and 19 men) with the diagnosis of an esophageal perforation have been treated at the Klinikum Oldenburg. The mean age at diagnosis was 66.4 years (range: 16–91 years). The data were retrospectively evaluated (Table 1). Results Etiology, Localization, And Diagnostics
Regarding the etiology of esophageal perforation, four groups exist: (a) 18/39 iatrogenic/interventional-associated (2 transesophageal echocardiography (TEE), 1 cystic duct stent replacement, 1 endoscopic retrograde cholangio pancreatography (ERCP), 1 polypectomy of the duodenum, 1 submucosal dissection of Barrett’s esophagus, 6 dilatations, 3 diagnostic endoscopies), (b) 9/39 accidental/foreign body ingestion, (c) 4/39 Boerhaave’s Syndrome, and (d) 3/39 postoperative and 5 events could not be classified. A safe inquiry about the presenting symptoms could not be obtained from the available data. We refer to our initial analysis in which one-quarter of the patients presented with leading symptoms suggestive of esophageal perforation (7). The most common localization of an esophageal perforation was the distal third with 16/29, followed by the middle third 14/39, and the proximal third with 9/39 patients. An upper endoscopy and a computed tomography (CT) scan with contrast were performed in all patients. The mean size of the defect was 2.4 cm (range: 5 cm, n = 19; however, size of the defect was not documented in all patients). Treatment
Of the 39 patients, 17 were referred to us from the surrounding hospital which partially performed frustrating therapeutic attempts, 9 patients presented to our surgical emergency room, and 13 patients had already been treated internally. Treatment was initiated within 12 h of the event in 18/39 patients, at >12 h in 5/39 patients, and at >24 h in 16/39 patients (Table 2). Conservative/interventional treatment A total of 22/39 patients were treated conservatively or interventional. Of the 22 patients, 7 were treated by stenting, 1 patient was treated by application of hemoclips, and 14 were treated by observation and/ or recurrent endoscopic lavages. In five patients, an interventional procedure was followed by surgery. Out of those patients, two received surgery, as we were unable to resect (R0) the adenocarcinoma endoscopically. Thus, resection of the malignancy was
performed after primary stabilization with a stent. In n = 1, the state of the patient deteriorated after initial stenting, leading to surgery. In n = 2, initial stenting with subsequent surgery was the treatment plan. Insertion of a thoracic drainage was not performed routinely in the interventional group, rather dependent on the CT result. Surgical treatment In n = 17 patients, an initial surgical procedure was performed. Different surgical techniques were used. In n = 5, stitching of the defect was performed; in n = 3, stitching of the defect with subsequent fundoplicatio was performed, of which n = 1 received interposition of the colon in a second setting, n = 1 received esophagectomy with gastric interposition in a second setting, n = 1 received esophagectomy with gastric interposition; n = 3 received esophagectomy with colon interposition; n = 2 received discontinuity resection of the proximal stomach and distal esophagus without reconstruction; in n = 1, abdominal approach through a right laparotomy, transhiatal lavage, and drainage were performed; and in n = 1, a collar approach with mediastinotomy and drainage was performed. The latter approach was used in treating a defect of the proximal esophagus after the extraction of glass fragments. In n = 3 patients who received an interposition of the colon, a gastric interposition was not performed because in n = 1, acid ingestions injured the stomach and in n = 1, gastric/esophageal gangrene occurred post fundoplicatio. Outcome
A total of n = 31 survived after esophageal perforation and n = 8 passed away, correlating to a mortality of about 21%. While looking at the patient collective according to the time passed before treatment was initiated, one notices that the survival rate was at 94% within the early treatment group (12 h). This outcome correlates with available published literature. Out of n = 8 who passed away, n = 1 was unable to receive surgical treatment due to her reduced state of health. Another patient was referred to us with a fulminant sepsis who deceased shortly after. Another patient passed away due to an endocarditis and a pulmonary embolism secondary to Boerhaave’s Syndrome. N = 2 deceased postoperatively due to a reduced state of health although no clear focus of infection could be identified. One died spontaneously due to an underlying hematothorax secondary to a pulmonary artery bleeding. One patient passed away due to malignant cardiac arrhythmia. Discussion This retrospective analysis was done to characterize and describe patients with an esophageal perforation, who in part were referred, diagnosed, and received their initial treatment from surrounding hospitals. Due to its etiology and acute clinical presentation, current data from random-control studies do not exist.
66 79 16 48 50 75 74 80 86 86 64
80 65 62 60 68 56 91
64 82 85
72 69 89 64
77 64
81 60
85 54 44 49 58 65
61
53 47
59
1 2 3 4 5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21
22 23 24 25
26 27
28 29
30 31 32 33 34 35
36
37 38
39
m
w w
w
w w m m m m
w m
w w
m w w m
w w w
w m m w m m w
m m m m m w w w m w m
Sex
Middle
Distal Middle
distal
Proximal cardia Middle Proximal Distal Proximal
Middle Middle
Middle Middle
Proximal Distal Middle Distal
Cardia Cardia middle
Proximal Cardia Middle Proximal Distal Distal Proximal
proximal Middle Middle Distal Distal Distal Middle Proximal Distal Middle Complete
Defect localisation
0,5 cm 1)0,2 cm 2)0,3 cm 0,6 cm
n.k.
n.k. 1)1,5cm, 2) 2cm n.k. 1,5 cm o.a. 0,8 cm n.k. n.k.
1 cm n.k.
n.k. n.k. n.k. n.k.
n.k. n.k.
