Em .I Cardio-thorac
Surg (1990) 4: 211-213
Oesophageal resection after instrumental perforation S. C. Griffin, J. Desai, E. R. Townsend, and S. W. Fountain Department
of Thoracic Surgery, Haretield Hospital,
Harefield,
Middlesex.
Abstract. Between 1981 and 1987, 11 patients underwent oesophageal resection following endoscopic perforation of the oesophagus. They had a median age of 67 years with a range of 36 to 88 years. They all were managed either by one- or two-stage oesophageal resections. Six patients were perforated at other hospitals, 5 on site. Seven had carcinoma of the middle or lower third of the oesophagus, 2 of these were perforated at attempted palliative intubation. Four had benign peptic strictures perforated during dilatation. Seven were resected within the first 24 hours and 4 between 2 and 10 days after perforation. All 4 patients with benign disease survived but 4 of the 7 patients with cancer died giving an operative mortality of 36.3%. Respiratory complications were the most common postoperative problem: all deaths were attributed to respiratory failure. Perforation of the thoracic oesophagus carries a high mortality. Resection after perforation in benign strictures may be life saving in a potentially lethal condition but resection after perforation, even in operable cancer, still carries a high mortality. [Eur J Cardio-thorac Surg (1990) 4: 211-2131 Key words: Endoscopic oesophageal perforation - Resection - Mortality
Endoscopic perforation of the oesophagus presents a formidable clinical problem to all surgeons. There seems to be no consensus of opinion as to the most effective form of treatment for this often fatal condition. The therapeutic options include drainage [7], repair [l], exclusion [17], resection [13] and particularly for cervical perforations, medical treatment alone [I 81.We have reviewed the experience at this centre over the years 1981 to 1987 when no common treatment policy was pursued. In this retrospective study, we have attempted to evaluate one particular treatment option, namely resection. Received for publication: Accepted for publication:
September 26, 1989 December 4, 1989
UK
Patients and methods Between 1981 and 1987, 2067 oesophagoscopies were performed at Haretield Hospital. It is our practice to perform endoscopic examinations under a general anaesthetic. During this time, 30 endoscopic perforations were treated. Nineteen patients were perforated at this hospital giving a perforation rate of 0.9%. As a regional referral centre, a further 11 patients were transferred from district hospitals with instrumental oesophageal perforations. Of the 30 perforations treated, 6 were of the cervical oesophagus and 24 of the thoracic oesophagus. Table 1 summarizes the treatment policy and outcome of all oesophageal perforations treated over the period of this review. The cervical perforations were all either treated by surgical drainage of the abscess in the neck or a regimen of antibiotic therapy, both in association with non-oral nutrition. In the 6 cases treated, these regimes were successful. Two patients with perforations of the thoracic oesophagus that were treated by drainage alone with no attempt at repair or excision both died. The 11 patients that underwent oesophageal resection as primary treatment for endoscopic perforation are reviewed in more detail. Open repair of the perforation was not considered to be a clinical option as these patients either had oesophageal malignancy or complex and very tight distal stenoses. The median age of the 3 1 resected patients was 68 with a range of 36-88 years. There were 6 men and 5 women. Of the 11, 5 were perforated at this hospital and the remaining 6 referred from elsewhere. Two patients who were considered unsuitable for an elective resection of malignant strictures were perforated during attempted pulsion intubation of these strictures, I patient was perforated during bouginage. This patient was being dilated with gum-elastic bougies before the unit had access to mercury filled Maloney bougies.
Table 1. Relationship with perforations
between site and management
in all patients
Management
Cervical number/deaths
Thoracic number/deaths
Conservative Drainage Open repair Resection
3/O 3/O 0 0
6/O 2/2 5/O 11/4
TOotal
6/O
24,‘6
212 Table 2. Relationship between mode of perforation patients undergoing resection
and outcome in
Mode of perforation
Benign number/deaths
Malignant number/deaths
Bougie Instrument Intubation
IlO 3/O O/O
O/O 512 212
Table 3. Timing of resection come Time of perforation Within 24 h Within 48 h At 96 h At 10 days Total
after perforation
and eventual
out-
Number/Deaths 712 211 131 l/O 11/4
Seven patients had oesophageal tumours of the middle or lower third. Of these lesions, 6 were adenocarcinomas and 1 a squamous tumour. Four patients had complex benign peptic strictures that had in the past required repeated dilatations (Table 2). The timing of the resections after initial perforation is shown in Table 3. The timing of the resections to a certain extent reflected the institution at which the perforation had taken place: patients perforated at this hospital tended to be operated upon sooner than those referred from elsewhere. All patients were treated by resection of the perforated lesion substituting stomach for the resected oesophagus with the oesophagogastric anastomosis in the chest. The choice of the two-stage or the one-stage approach depended principally on the site of perforation but the patients with malignancies had two stage resections. One anastomosis was fashioned with circular stapler and the remainder were hand sewn in either 1 layer (6 patients) or 2 layers (5 patients) using interrupted absorbable sutures (3/O Vicryl).
Results Of the 11 patients resected, there were 4 deaths giving an in-patient mortality of 36%. There were no deaths in the group of 4 patients that were operated upon for benign disease. Respiratory failure was the major cause of morbidity and mortality in this series. Seven required prolonged ventilation. Of these, 4 underwent elective tracheostomy to facilitate weaning from ventilatory support. Two patients who had tracheostomies eventually succumbed to pneumonia, 2 more developed additional renal failure from which both died. All 4 deaths were attributable to respiratory failure. There was one stroke and one anastomotic leak from which the patients survived. The median postoperative stay for survivors was 21 days with a range of 12 to 34 days. The median time to death for the remainder was 14 days with a range of 9 to 60 days. Of the 7 patients operated upon within the first 24 h, 2 died. Two of the 4 patients operated upon at more than 24 h died.
Discussion The importance of early diagnosis of endoscopic perforation is widely acknowledged but there remains no consensus of opinion on the subsequent management of the problem. Oesophageal resection after instrumental perforation was first reported by Nealon in 1961 [13]. Since that time, the procedure has gained acceptance in certain clinical situations. Hendren resected 5 patients with benign disease with no deaths, 3 of the 5 were children with congenital oesophageal abnormalities [6]. Grill0 resected 8 patients after instrumental perforation. In 3 cases, oesophago-gastric continuity was achieved using colon. One patient in his study with adenocarcinoma died [l 11. Bladergroen resected 10 patients, 5 of whom were operated upon within the first 24 h and survived. He stressed the importance of early diagnosis and treatment [3]. Moghissi in his series of 39 patients treated after endoscopic perforation, resected Il. This was achieved with a mortality of 18% which he found contrasted with a mortality of 83% in the group with pre-existing strictures who were treated with less radical surgery [12]. During the period studied, this centre performed 206 elective oesophageal resections for carcinoma with an in-hospital mortality of 10.2% [5]. In this series, all perforations were in the presence of distal obstruction. The fundamental principles of surgical practice suggest that fistulae do not heal under these conditions. The aim of the treatment must be to either remove the perforation or repair it in combination with relief of the distal obstruction. In the case of an operable tumour, resection is considered by some the procedure of choice [4] and recently the use of subtotal oesophagectomy with cervical anastomosis following perforation has been advocated both to avoid an anastomosis in a contaminated field and to lessen the risk of reflux [lo]. Conventional resection is not generally thought to be the optimum treatment for peptic oesophageal strictures as the restenosis rate is unacceptably high [2]. Oesophageal lengthening techniques combined with an anti-reflux procedure are now considered to offer the best results in elective cases [15, 161. However after perforation, resection is the only option. The timing of resection after perforation is thought by some to have an influence on the eventual outcome. Some authors suggest that early treatment is advantageous and advocate medical management if the diagnosis is delayed [3, 141. Although the numbers in this series are small, our results tend to support this view. The problem of endoscopic oesophageal perforation is probably under reported and resection in this situation is similarly not widely reported. In this context, conclusions are difficult to draw, particularly from a retrospective series of such a small number of patients. However, we can state that emergency oesophageal resection at whatever time after instrumental perforation still carries a high mortality. Resection for benign disease is safe and when an operable tumour is perforated, the best course of action is to proceed to resection immediately before mediastinal contamination has occurred. Resection in patients who have already been declared unsuitable for surgery carries an unacceptable mortality.
213 References 1. Barret NR (1947) Report
2.
3.
4.
5.
6.
7. 8.
9. 10. 11.
of the case of spontaneous perforation of the oesophagus successfully treated by operation. Br J Surg 35:216-218 Belsey R (1965) Reconstruction of the oesophagus with left colon. J Thorac Cardiovasc Surg 49: 33-55 Bladergroen MR, Lowe JE, Postlewaite RW (1986) Diagnosis and recommended management of oesophageal perforation and rupture. Ann Thorac Surg 42:235-239 Brewer LA, Carter R, Mulder GA, Stiles QR (1986) Options in the management of perforation of the oesophagus. Am J Surg 152.64-69 Griffin SC, Desai J. Charlton M. Townsend ER, Fountain SW (1989) Factors influencing morbidity and mortality following oesophageal resection. Eur J Cardio-thorac Surg 3:419-424 Hendren WH, Bruce M, Henderson MB (1968) Immediate esophagectomy for instrumental perforation of the thoracic esophagus. Ann Surg 166:998&1003 Keen G (1968) The surgical management of old oesophageal perforation. J Thorac Cardiovasc Surg 56:603-606 Lewis I (1947) The surgical treatment of carcinoma of the oesophagus, with a special reference to a new operation for growth of the middle third. Br J Surg 34: 18-30 Logan A (1963) The surgical treatment of carcinoma of the oesophagus and cardia. J Thorac Cardiovasc Surg 46: 150- 161 Mathews HR. Mitchel IM, McGuigan JA (1989) Emergency subtotal oesophagectomy. Br J Surg 76:918-920 Michel L, Grill0 H, Malt RA (1981) Operative and non-operative management of esophageal perforations. Ann Surg 194:58 -63
12. Mohgissi K, Pender D (1988) Instrumental perforation of the oesophagus and their management. Thorax 43: 642-646 13. Nealon TF, Templeton JY. Cuddy VD. Gibbon JH (1961) Instrumental perforation of the oesophagus. J Thorac Cardiovasc Surg 41: 75-77 14. Nesbitt JC, Sawyers JL (1987) Surgical management of esophageal peforation. Am Surg 53: 183-191 15. Orringer MD, Sloan H (1978) Combined Collis-Nissen reconstruction of the oesophagogastric junction. Ann Thorac Surg 25: 16-21 16. Pearson FG, Langer B, Henderson RD (1971) Gastroplasty and Belsey hiatus hernia repair. J Thorac Cardiovasc Surg 61: 50-63 17. Urschel HC, Razzuk MA, Wood RE, Galbrath N, Polkey N. Paulson D (1974) Improved management of oesophageal perforation; exclusion and diversion in continuity. Ann Surg 179: 587-591 18. Wesdorf ICE, Bartelman JFWM, Huibregtse K, Den Hartog FCA, Tytgat GN (1984) Treatment of instrumental oesophageal perforation. Gut 25: 398-404
Mr. S. W. Fountain, FRCS Consultant Thoracic Surgeon Harefield Hospital Haretield Middlesex UB9 6JH UK
Surg (1990) 4: 211-213
Oesophageal resection after instrumental perforation S. C. Griffin, J. Desai, E. R. Townsend, and S. W. Fountain Department
of Thoracic Surgery, Haretield Hospital,
Harefield,
Middlesex.
Abstract. Between 1981 and 1987, 11 patients underwent oesophageal resection following endoscopic perforation of the oesophagus. They had a median age of 67 years with a range of 36 to 88 years. They all were managed either by one- or two-stage oesophageal resections. Six patients were perforated at other hospitals, 5 on site. Seven had carcinoma of the middle or lower third of the oesophagus, 2 of these were perforated at attempted palliative intubation. Four had benign peptic strictures perforated during dilatation. Seven were resected within the first 24 hours and 4 between 2 and 10 days after perforation. All 4 patients with benign disease survived but 4 of the 7 patients with cancer died giving an operative mortality of 36.3%. Respiratory complications were the most common postoperative problem: all deaths were attributed to respiratory failure. Perforation of the thoracic oesophagus carries a high mortality. Resection after perforation in benign strictures may be life saving in a potentially lethal condition but resection after perforation, even in operable cancer, still carries a high mortality. [Eur J Cardio-thorac Surg (1990) 4: 211-2131 Key words: Endoscopic oesophageal perforation - Resection - Mortality
Endoscopic perforation of the oesophagus presents a formidable clinical problem to all surgeons. There seems to be no consensus of opinion as to the most effective form of treatment for this often fatal condition. The therapeutic options include drainage [7], repair [l], exclusion [17], resection [13] and particularly for cervical perforations, medical treatment alone [I 81.We have reviewed the experience at this centre over the years 1981 to 1987 when no common treatment policy was pursued. In this retrospective study, we have attempted to evaluate one particular treatment option, namely resection. Received for publication: Accepted for publication:
September 26, 1989 December 4, 1989
UK
Patients and methods Between 1981 and 1987, 2067 oesophagoscopies were performed at Haretield Hospital. It is our practice to perform endoscopic examinations under a general anaesthetic. During this time, 30 endoscopic perforations were treated. Nineteen patients were perforated at this hospital giving a perforation rate of 0.9%. As a regional referral centre, a further 11 patients were transferred from district hospitals with instrumental oesophageal perforations. Of the 30 perforations treated, 6 were of the cervical oesophagus and 24 of the thoracic oesophagus. Table 1 summarizes the treatment policy and outcome of all oesophageal perforations treated over the period of this review. The cervical perforations were all either treated by surgical drainage of the abscess in the neck or a regimen of antibiotic therapy, both in association with non-oral nutrition. In the 6 cases treated, these regimes were successful. Two patients with perforations of the thoracic oesophagus that were treated by drainage alone with no attempt at repair or excision both died. The 11 patients that underwent oesophageal resection as primary treatment for endoscopic perforation are reviewed in more detail. Open repair of the perforation was not considered to be a clinical option as these patients either had oesophageal malignancy or complex and very tight distal stenoses. The median age of the 3 1 resected patients was 68 with a range of 36-88 years. There were 6 men and 5 women. Of the 11, 5 were perforated at this hospital and the remaining 6 referred from elsewhere. Two patients who were considered unsuitable for an elective resection of malignant strictures were perforated during attempted pulsion intubation of these strictures, I patient was perforated during bouginage. This patient was being dilated with gum-elastic bougies before the unit had access to mercury filled Maloney bougies.
Table 1. Relationship with perforations
between site and management
in all patients
Management
Cervical number/deaths
Thoracic number/deaths
Conservative Drainage Open repair Resection
3/O 3/O 0 0
6/O 2/2 5/O 11/4
TOotal
6/O
24,‘6
212 Table 2. Relationship between mode of perforation patients undergoing resection
and outcome in
Mode of perforation
Benign number/deaths
Malignant number/deaths
Bougie Instrument Intubation
IlO 3/O O/O
O/O 512 212
Table 3. Timing of resection come Time of perforation Within 24 h Within 48 h At 96 h At 10 days Total
after perforation
and eventual
out-
Number/Deaths 712 211 131 l/O 11/4
Seven patients had oesophageal tumours of the middle or lower third. Of these lesions, 6 were adenocarcinomas and 1 a squamous tumour. Four patients had complex benign peptic strictures that had in the past required repeated dilatations (Table 2). The timing of the resections after initial perforation is shown in Table 3. The timing of the resections to a certain extent reflected the institution at which the perforation had taken place: patients perforated at this hospital tended to be operated upon sooner than those referred from elsewhere. All patients were treated by resection of the perforated lesion substituting stomach for the resected oesophagus with the oesophagogastric anastomosis in the chest. The choice of the two-stage or the one-stage approach depended principally on the site of perforation but the patients with malignancies had two stage resections. One anastomosis was fashioned with circular stapler and the remainder were hand sewn in either 1 layer (6 patients) or 2 layers (5 patients) using interrupted absorbable sutures (3/O Vicryl).
Results Of the 11 patients resected, there were 4 deaths giving an in-patient mortality of 36%. There were no deaths in the group of 4 patients that were operated upon for benign disease. Respiratory failure was the major cause of morbidity and mortality in this series. Seven required prolonged ventilation. Of these, 4 underwent elective tracheostomy to facilitate weaning from ventilatory support. Two patients who had tracheostomies eventually succumbed to pneumonia, 2 more developed additional renal failure from which both died. All 4 deaths were attributable to respiratory failure. There was one stroke and one anastomotic leak from which the patients survived. The median postoperative stay for survivors was 21 days with a range of 12 to 34 days. The median time to death for the remainder was 14 days with a range of 9 to 60 days. Of the 7 patients operated upon within the first 24 h, 2 died. Two of the 4 patients operated upon at more than 24 h died.
Discussion The importance of early diagnosis of endoscopic perforation is widely acknowledged but there remains no consensus of opinion on the subsequent management of the problem. Oesophageal resection after instrumental perforation was first reported by Nealon in 1961 [13]. Since that time, the procedure has gained acceptance in certain clinical situations. Hendren resected 5 patients with benign disease with no deaths, 3 of the 5 were children with congenital oesophageal abnormalities [6]. Grill0 resected 8 patients after instrumental perforation. In 3 cases, oesophago-gastric continuity was achieved using colon. One patient in his study with adenocarcinoma died [l 11. Bladergroen resected 10 patients, 5 of whom were operated upon within the first 24 h and survived. He stressed the importance of early diagnosis and treatment [3]. Moghissi in his series of 39 patients treated after endoscopic perforation, resected Il. This was achieved with a mortality of 18% which he found contrasted with a mortality of 83% in the group with pre-existing strictures who were treated with less radical surgery [12]. During the period studied, this centre performed 206 elective oesophageal resections for carcinoma with an in-hospital mortality of 10.2% [5]. In this series, all perforations were in the presence of distal obstruction. The fundamental principles of surgical practice suggest that fistulae do not heal under these conditions. The aim of the treatment must be to either remove the perforation or repair it in combination with relief of the distal obstruction. In the case of an operable tumour, resection is considered by some the procedure of choice [4] and recently the use of subtotal oesophagectomy with cervical anastomosis following perforation has been advocated both to avoid an anastomosis in a contaminated field and to lessen the risk of reflux [lo]. Conventional resection is not generally thought to be the optimum treatment for peptic oesophageal strictures as the restenosis rate is unacceptably high [2]. Oesophageal lengthening techniques combined with an anti-reflux procedure are now considered to offer the best results in elective cases [15, 161. However after perforation, resection is the only option. The timing of resection after perforation is thought by some to have an influence on the eventual outcome. Some authors suggest that early treatment is advantageous and advocate medical management if the diagnosis is delayed [3, 141. Although the numbers in this series are small, our results tend to support this view. The problem of endoscopic oesophageal perforation is probably under reported and resection in this situation is similarly not widely reported. In this context, conclusions are difficult to draw, particularly from a retrospective series of such a small number of patients. However, we can state that emergency oesophageal resection at whatever time after instrumental perforation still carries a high mortality. Resection for benign disease is safe and when an operable tumour is perforated, the best course of action is to proceed to resection immediately before mediastinal contamination has occurred. Resection in patients who have already been declared unsuitable for surgery carries an unacceptable mortality.
213 References 1. Barret NR (1947) Report
2.
3.
4.
5.
6.
7. 8.
9. 10. 11.
of the case of spontaneous perforation of the oesophagus successfully treated by operation. Br J Surg 35:216-218 Belsey R (1965) Reconstruction of the oesophagus with left colon. J Thorac Cardiovasc Surg 49: 33-55 Bladergroen MR, Lowe JE, Postlewaite RW (1986) Diagnosis and recommended management of oesophageal perforation and rupture. Ann Thorac Surg 42:235-239 Brewer LA, Carter R, Mulder GA, Stiles QR (1986) Options in the management of perforation of the oesophagus. Am J Surg 152.64-69 Griffin SC, Desai J. Charlton M. Townsend ER, Fountain SW (1989) Factors influencing morbidity and mortality following oesophageal resection. Eur J Cardio-thorac Surg 3:419-424 Hendren WH, Bruce M, Henderson MB (1968) Immediate esophagectomy for instrumental perforation of the thoracic esophagus. Ann Surg 166:998&1003 Keen G (1968) The surgical management of old oesophageal perforation. J Thorac Cardiovasc Surg 56:603-606 Lewis I (1947) The surgical treatment of carcinoma of the oesophagus, with a special reference to a new operation for growth of the middle third. Br J Surg 34: 18-30 Logan A (1963) The surgical treatment of carcinoma of the oesophagus and cardia. J Thorac Cardiovasc Surg 46: 150- 161 Mathews HR. Mitchel IM, McGuigan JA (1989) Emergency subtotal oesophagectomy. Br J Surg 76:918-920 Michel L, Grill0 H, Malt RA (1981) Operative and non-operative management of esophageal perforations. Ann Surg 194:58 -63
12. Mohgissi K, Pender D (1988) Instrumental perforation of the oesophagus and their management. Thorax 43: 642-646 13. Nealon TF, Templeton JY. Cuddy VD. Gibbon JH (1961) Instrumental perforation of the oesophagus. J Thorac Cardiovasc Surg 41: 75-77 14. Nesbitt JC, Sawyers JL (1987) Surgical management of esophageal peforation. Am Surg 53: 183-191 15. Orringer MD, Sloan H (1978) Combined Collis-Nissen reconstruction of the oesophagogastric junction. Ann Thorac Surg 25: 16-21 16. Pearson FG, Langer B, Henderson RD (1971) Gastroplasty and Belsey hiatus hernia repair. J Thorac Cardiovasc Surg 61: 50-63 17. Urschel HC, Razzuk MA, Wood RE, Galbrath N, Polkey N. Paulson D (1974) Improved management of oesophageal perforation; exclusion and diversion in continuity. Ann Surg 179: 587-591 18. Wesdorf ICE, Bartelman JFWM, Huibregtse K, Den Hartog FCA, Tytgat GN (1984) Treatment of instrumental oesophageal perforation. Gut 25: 398-404
Mr. S. W. Fountain, FRCS Consultant Thoracic Surgeon Harefield Hospital Haretield Middlesex UB9 6JH UK
