Br. J. Surg. Vol. 62 (1975) 448-454
The management of peptic oesophageal stricture R. C. N. W I L L I A M S O N * SUMMARY
Eighty-three adult pafients with peptic oesophageal stricture are reviewed with regard to diagnosis and treatment. Sliding hiatus hernia was present in 94 per cent of cases. The strictures were graded according to the degree of stenosis encountered at initial oesophagoscopy, but several othzr factors influenced the management of each case. Bouginage should always be attempted before proczeding to surgery. Thirty-eight patients were treated by bouginage alone, with 21 (55 per cent) good results and 3 deaths. Forty-five patients came to surgery, with 32 (71 per cent) good results and 3 deaths. Six patients sustained instrumental perforation of the oesophagus with I fatality. There was in addition 1 case of silent perforation following selfbouginage, which is believed 10 be unique. Hiatal herniorrhaphy combined with simple bouginage is recommended where reflux oesophagitis dominates the clinical picture. Local plastic repair (oesophagoplasty) has proved a simple and effective method of treating low annular strictures. Oesophagogastrostomy was carried out in 12 patients without operative loss and with good results in 8 cases; the addition of a pyloroplasty is believed to lessen postoperative reflux.
INJURYto the oesophagus may cause submucosal scarring, leading to permanent stenosis. In this country the commonest injurious agent is gastro-oesophageal reflux, but treatment of the resultant peptic stricture remains controversial. It has been stated that ‘bouginage should be the method of choice in every case unless one is dealing with severe haemorrhage or perforation’ (Benedict, 1966). Most British surgeons, however, advocate operation for the majority of cases (Editorial, 1968), though this may vary from local plastic repair to extensive resection with oesophageal replacement. The present communication reviews a large series of patients with peptic oesophageal stricture, with particular regard to their diagnosis and treatment in a district thoracic surgical unit.
Patients The series comprises 83 patients with peptic stricture treated at Peppard Hospital, Oxfordshire, during the Table I: AETIOLOGICAL TYPES OF BENIGN OESOPHAGEAL STRICTURE Aetiological type: Peptic NO. Of Cases: 83 Anastomotic 8 Sideropenic 2 Pemphigoid 1 Irradiation 1 Corrosive 1 -
Totdl
448
96
Fig. 1. Barium swallow film showing an early annular stricture of the lower oesophagus. This caused surprisingly severe symptoms but was successfully treated by local plastic repair (oesophagopldsty).
10-year period 1962-71 under the care of one consultant thoracic surgeon. Table I shows that this was by far the commonest aetiological type of benign oesophageal stricture seen during this period. All the patients were followed up for a minimum of 2 years after treatment. The total includes 49 women and 34 men, aged from 14 to 89 years with a mean of 68.2 years. The maximal incidence occurred during the seventh and eighth decades, two-thirds of patients being over the age of 70. Not surprisingly this elderly population had a high incidence of cardiorespiratory disease, 1 I patients being severely incapacitated.
* Royal Berkshire Hospital, Reading. Present address: United Bristol Hospitals.
Peptic oesophageal stricture
a
b
Fig. 2. a , ‘Ascending oesophageal fibrosis’. Barium swallow film showing an elongated peptic stricture. b, The appearance on a barium swallow film following resection of the stricture with jejunal replacement. The patient obtained an excellent result.
Sliding hiatus hernia was demonstrated in 78 of the 83 cases (94 per cent), while the remaining 5 patients all had clinical and endoscopic features of reflux oesophagitis. Three patients with no antecedent gastro-oesophageal symptoms developed an acute postoperative stricture, in each case following colonic surgery. Diagnosis Clinical Table II shows that dysphagia was the only constant symptom. It varied greatly in severity and duration and 7 patients presented with complete oesophageal obstruction (aphagia). It is notable that one-third of the patients had no previous history of reflux oesophagitis. Radiological Apart from 2 patients with aphagia who underwent emergency oesophagoscopy, all the strictures were first demonstrated by barium swallow examination. Figs. I and 2 illustrate the two classic types, namely annular stricture caused by a localized ring of fibrous tissue, and ascending oesophageal fibrosis (Kelly, 1936) where scarring extends for several centimetres with muscular hypertrophy, peri-oesophagitis and rnediastinal adenopathy. In 12 strictures subsequently shown to be benign there was radiological suspicion of
Fig. 3. Barium swallow film magnified to show a rounded filling defect in the lower oesophagus. This was reported as a leiomyoma but at thoracotomy the ‘tumour’ proved to be a peanut, swollen and impacted above an annular stricture. Oesophagoplasty was successfully performed. Table 11: SYMPTOMS OF PEPTIC STRICTURE No. of Symptom cases ;4 Dysphagia 83 100 Reflux oesophagitis (heartburn, 55 66 acid brash etc.) Regurgitation of solids 42 51 Weight loss (>,6,3 kg) 19 23 Haematemesis 8 9 Inhalation pneumonitis 5 6 Hiccuo 5 6
carcinoma; 5 of these also appeared irregular at endoscopy so that the possibility of carcinoma became a major factor in advising surgery. Fig. 3 demonstrates another pitfall in radiological diagnosis. Endoscopic In nearly every case the presence of a stricture was confirmed at oesophagoscopy and definitive treatment by bouginage was attempted at the same time. Some degree of oesophagitis was observed in 45 patients (54 per cent). Fig. 4 shows that the large majority of peptic strictures were confined to the lower third of the oesophagus, as expected. The strictures were graded according to the degree of stenosis encountered at the initial oesophagoscopy as follows: Grade I : Slight stenosis easily negotiated. Grade I I : Moderate stricture dilated without difficulty. Grade 111: Tight strictureonly dilated withdifficulty. Grade I V : Impassable stricture. 449
R. C. N. Williamson Distance from incisor teeth (cm)
Peptic strictures
Non-peptic strictures
ss
10
. .. .......... .... ..... ............ ...... ...... .... .......
Pg
R
tc A AAA
A AA A S=Sideropenic Pg= Pemphigoid R= Irradiation C=Corrosive A= Anastomotic
Fig. 4. Diagram to show the level of benign oesophageal strictures at endoscopy. Each peptic stricture is marked by a black circle. Strictures of non-peptic origin are included for comparison. The highest peptic stricture was found a t 18 cm from the incisor teeth; it proved amenable to bouginage.
Histologica I In 53 patients carcinoma was excluded as far as possible by the examination of biopsy or operative specimens, and in another 13 patients the oesophagus was directly inspected at thoracotomy. In the remaining 17 cases the diagnosis rested upon the continued absence of suspicious clinical, radiological or endoscopic features at follow-up. There was no incidence of a benign peptic stricture undergoing malignant change.
Methods of treatment The primary objective of treatment was to enable every patient to swallow sufficiently well to maintain his weight and nutritional status, albeit with some modification of diet. Associated oesophagitis was treated by a standard medical regime (Bennett, 1973). Dilatation of strictures was carried out using Neoplex or metal Souttar bougies, usually under general anaesthesia and always under direct endoscopic control, though 1 I patients whose stricture demanded repeated dilatation at frequent intervals were taught to pass their own bougies under a local anaesthetic. Fibre-optic oesophagoscopy was not available during the period of this study. The following surgical methods were employed. 450
Repair of hiatus hernia entailed exposure of the oesophagus within the chest, usually via a left thoracolaparotomy. Reduction of the hernia was achieved as far as possible and the hiatus was repaired by anterior apposition of the crura; routine HeinekeMikulicz pyloroplasty was carried out additionally. Operative dilatation of awkward low strictures was performed via a gastrotomy to allow digital dilatation or the retrograde passage of bougies. In a few such cases a Celestin tube was passed from above, right through the stricture and out, to ensure that adequate dilatation had been achieved. In certain selected cases the tube was left in situ. Oesophagoplasty or local plastic repair was reserved for low annular strictures which were exposed within the chest, incised longitudinally through all coats and closed transversely with one layer of interrupted catgut sutures, in a manner akin to Heineke pyloroplasty. Resection of the stricture was generally performed through a left thoracolaparotomy with direct oesophagogastric anastomosis and the addition of a pyloroplasty. Oblique transection of the stomach allowed the creation of a gastric tube for anastomosis with the transected oesophagus. In 3 cases a segment ofjejunum or right colon was brought up to bridge the defect following an extensive resection.
Peptic oesophageal stricture The long term result of treatment in each case was assessed as follows : Good: Manages virtually normal diet; minimal dysphagia. Moderate: Careful diet needed ; occasional regurgitation of food. Poor: Liquid diet; frequent dysphagia and regurgitation.
Where the result remained only ‘moderate’ after repeated bouginage surgery was advised unless contraindicated by extreme age o r infirmity, and these patients comprise the ‘bouginage ineffective’ group listed in Table IV. Forty-five patients were treated surgically (54 per cent) and all but 5 of these had undergone a previous trial of bouginage. Three types of surgical procedure were performed :
Results of treatment Of 55 patients presenting with symptoms of reflux, 24 continued to require medical treatment for this after either bouginage o r surgery. However, only 3 patients in all had symptoms of reflux aggravated by local treatment of the stricture. Table III shows that whatever the grading of the stricture at initial endoscopy roughly equal numbers of patients were treated by surgery and bouginage. Though grade I strictures responded readily to simple bouginage, surgery was often required in this group for severe associated reflux oesophagitis. The largest group of strictures belonged to grade 11; of these, half were treated surgically and half by bouginage, which usually became progressively easier. Grade I l l and IV strictures were very unpredictable. Five patients reported some improvement in swallowing despite unsuccessful bouginage, 2 of them requiring no further treatment. Five strictures, initially impassable, subsequently yielded t o bouginage, and the long term result in these difficult cases was sometimes surprisingly good. Nonetheless difficulties encountered during bouginage (including previous endoscopic perforation in 3 cases) were a common indication for surgery (Table I V ) if dysphagia subsequently recurred. The results of treatment are shown in Table V. Thirty-eight patients (46 per cent) were managed by bouginage alone, 17 requiring only one dilatation carried out at the time of diagnostic oesophagoscopy.
I . Repair of hiatus hernin A relatively young group of patients with marked oesophagitis was submitted to anti-reflux surgery with bouginage of the associated grade I o r 11 stricture. This combination was employed in 1 case of grade I11 stricture presenting in a girl of 14, but the result was poor and she required subsequent resection with oesophagogastrostomy. Table V shows that the results of this combined procedure were otherwise very Table 111: GRADING OF PEPTIC STRICTURES Grade
Bouainane
Surgery
II 44
5 22
9 11 8
4 6 1
6 22 5
83
38
I I1 111 1v Unclassified *
Total
Table IV: INDICATIONS FOR SURGERY Indication: Bouginage difficult N o. of cases: 14 Bouginage ineffective 13 12 Reflux oesophagitis Carcinoma suspected 5 ‘Leiomyoma’ (Fig. 3 ) 1 -
Total 45
Result Good
Moderate
Poor
Death
38 45 12 6 9 7 12 3
21 32
14 9 1 0 4 1 3
0 1 1
3 3 0 0 2 0 0
83
53 (64%)
10 5 3 6 8 2
0 23 (28%)
F F F F F F
1 0 0
1 0 1 ( 1%)
1 6 (7%)
one type of operative procedure was carried out and the final surgical result is recorded.
Table VI: CASES OF INSTRUMENTAL PERFORATION OF THE OESOPHAGUS Delay in onset of symptoms Level of Difficulty Age Sex at endoscopy (h) perforation Treatment 70 77 88 37 78 79
7
45
No. of cases
Total
* In 4 cases more than
5
* Stricture not seen at endoscopy or bouginage not attempted.
Table V: RESULTS OF TREATMENT Procedure Bouginage Surgery* Repair of hiatus hernia Retrograde dilatation Oesophageal intubation Oesophagoplasty Oesophagogastrostomy Jejunum/colon replacement
Treatment
N o. of cases
Postcricoid Postcricoid Nil Nil Grade 111 stricture Grade 111 stricture
24 24 24 24 4 2
Cervical Cervical Thoracic Thoracic Thoracic Thoracic
Conservative Conservative Conservative Conservative Operative Operative
Complications Nil Nil Empyema thordcis Peri-oesophageal abscess Nil Death
451
R. C. N. Williamson satisfactory within the period of follow-up, even though extensive peri-oesophagitis often prevented complete reduction of the hernia. The stricture was treated by prior endoscopic bouginage or, i n 2 cases, by retrograde digital dilatation at the time of operation. In 3 cases bouginage was required for a short time postoperatively as well. 2. Limited surgery to the stricture Operative dilatation was performed in 6 cases, including 1 man who underwent retrograde bouginage no less than three times in 4 months before obtaining a satisfactory and lasting result. The 1 patient with a ‘poor’ result had sustained a previous endoscopic perforation and was deemed too frail for any further procedure. Oesophageal intubation was reserved for a group of very elderly and unfit patients in whom bouginage was inadequate but major surgery impossible. It was not a particularly satisfactory procedure, for 2 patients died and several of the survivors had trouble with the tube. In 3 cases it slipped up the oesophagus to cause renewed dysphagia and require removal. One old lady passed her tube in two pieces per rectum to the consternation of the nursing staff. Oesophagoplasty was carried out in 7 cases of localized annular stricture close to the cardia (Figs. 1, 3). The results were good apart from 1 patient with residual dysphagia due to disordered oesophageal motility. Two patients with anastomotic stricture were treated with equal success by this procedure.
3 . Resection Oesophagogastrostomy was carried out in 12 cases; the oldest patient was 72 and all enjoyed reasonable general health. Although several had initial postoperative problems 8 patients obtained a ‘good’ result. Reflux oesophagitis spoilt the result in 3 cases, aggravated in one by obesity and in another by aspirin. Four patients needed limited postoperative bouginage. The only ‘poor’ result was in a patient who bled from an anastomotic ulcer, requiring further resection with jejunal interposition. This operation was performed in one other patient (Fig. 2) and neither case of jejunal replacement has required any further treatment. The sole case of colonic interposition died following an anastomotic leak on the tenth day, this being the only fatality among the 14 cases of resection. Complications Three patients died following endoscopic bouginage, a mortality of 7.9 per cent. The causes of death were pulmonary embolism, chronic chest infection and instrumental perforation. Two patients sustained a limited gastro-intestinal haemorrhage following instrumentation. There were 6 cases of instrumental perforation of the oesophagus during 270 endoscopies performed for benign stricture, an incidence of 2.2 per cent; the details are shown in Table VI. Strangely, 5 of the 6 patients had undergone previous uneventful 452
bouginage. In 2 cases dilatation was completed without suspicion of trauma, in 2 difficulty was encountered at the postcricoid level and in 2 the stricture itself was difficult to dilate. These last 2 both presented within a few hours of the accident and were treated surgically, though in I case operation was delayed for 48 hours and the patient died. The other 4 were treated conservatively with intravenous fluids and antibiotics in elixir form. Two of these developed a local abscess but made a complete recovery following drainage; the 2 patients with postcricoid perforation had minimal upset. The following case report illustrates one remarkable complication of self-bouginage : A 65-year-old man presented with dysphagia, epigastric pain and regurgitation of solids. Gross kyphosis was noted. Barium swallow examination showed a large hiatus hernia and an irregular stricture at the cardia with rigid margins suspicious of carcinoma. Endoscopy revealed an inflamed fibrous stricture at 35 cm which could not be dilated; biopsies showed no carcinoma. Ten days later the stricture was successfully dilated. The patient survived a major pulmonary embolus thereafter and his swallowing was much improved. In the following year the stricture was dilated three times and as it remained tight he was taught self-bouginage; this he practised on three occasions, the last time passing a no. 43 FG Neoplex bougie without difficulty. After this he began to develop abdominal pain and was readmitted with a small haematemesis 16 days after self-bouginage. Findings of peritonitis led to laparotomy and the discovery of a large left subhepatic abscess with a second smaller abscess within the liver. These were drained and the patient made an uneventful recovery. Two subsequent attempts a t endoscopic bouginage have failed but his swallowing is reasonably satisfactory.
Postoperatively, 3 patients died, a mortality of 6.7 per cent. Two deaths (from haematemesis and aspiration pneumonia respectively) followed insertion of a Celestin tube. The other fatality followed rupture of an interposed segment of colon. There was 1 case of postoperative chylothorax requiring thoracic duct ligation and one of strangulated diaphragmatic incisional hernia.
Discussion Reflux oesophagitis, the commonest disease affecting the gullet, may eventually lead to peptic stricture in 2-3 per cent of cases (Johnson and Lukash, 1970). Under certain circumstances (Allison, 1970), and especially after any major abdominal operation (Benedict and Gillespie, 1954), this process may develop rapidly; in the present series 3 patients developed an acute peptic stricture, in each case following colonic surgery. Benedict (1966) advocated bouginage for nearly every peptic stricture, reporting good results in 70 per cent of 133 patients. Many of these with unrepaired hiatal hernia became entirely symptom-free following bouginage. Barrett (1962) objected to simple bouginage on the grounds that it broke down a natural barrier against the upwards spread of oesophagitis, but in the present series reflux was only aggravated in 3 patients undergoing either bouginage or limited surgery to the stricture. Nevertheless, few surgeons would agree with Benedict that haemorrhage and perforation comprise the only indications for surgery. Only 55 per cent of
Peptic oesophageal stricture our patients obtained a good result from bouginage alone, and its failure to procure lasting relief from dysphagia was a common indication for operation. Likewise, difficulty encountered during previous bouginage and the possibility of carcinoma will usually demand surgical intervention. There were 6 cases of instrumental perforation during endoscopy, an incidence of 2.2 per cent, with 1 fatality. Raptis and Milne (1972) reported 4 such cases in their series of 100 benign strictures, again with 1 fatality. The overall incidence of perforation during all upper gastro-intestinal endoscopies is 0.5-0.7 per cent (Elner and Dahlback, 1962; Mark and Knauer, 1969), occurring most frequently in the cervical oesophagus in elderly women (Wychulis et al., 1969). This accident, of which 2 examples are reported here, can usually be managed conservatively (Hardin et al., 1967). Dilatation for benign or malignant stricture is the second commonest cause of iatrogenic perforation (Loop and Groves, 1970). The evidence suggests that patients who present with features of ruptured thoracic oesophagus within a few hours of endoscopy require immediate surgical exploration. The case of silent perforation during self-bouginage is believed to be without precedent, although Goldstein and Sherlock (1966) reported 1 case of unsuspected perforation following endoscopic bouginage carried out under a local anaesthetic. Endoscopic bouginage is therefore not without risk and in the present series carried a mortality (7.9 per cent) akin to that of operation (6.7 per cent). Furthermore, the 3 postoperative deaths occurred in patients in whom surgery had been necessitated by failure of bouginage. Taken in isolation, the tightness of the stricture was found to be a poor guide in determining the need for surgery. The natural history of peptic stricture appears to be unpredictable and the degree of reflux oesophagitis varies widely between cases. Age and infirmity often preclude the adoption of more vigorous methods of treatment, though it may be a matter of judgement whether multiple bouginage or limited surgery carries a greater risk in such cases. Twenty-one per cent of all cases required no other treatment than a single dilatation of the stricture carried out at the time of diagnostic oesophagoscopy. The policy was therefore to attempt bouginage in nearly every case, but if this should fail, to proceed to surgery, the choice of operation depending upon the degree of oesophagitis and the extent of the stricture. Where symptoms of reflux overshadow those of stricture lasting relief from dysphagia can usually be achieved by combining hiatal repair with simple bouginage (Hill et al., 1970). Palmer (1968) reported 7 patients treated in this way who subsequently required bouginage as often as those whose hernia had been left unrepaired, but in the present series this was not the case: of I 1 patients with early or moderate grades of stricture submitted to hiatal herniorrhaphy, only 3 needed any postoperative bouginage at all and that only for a brief period. This was in spite of the fact that gross peri-oesophagitis often prevented complete reduction of the hernia. 33
Where the stricture itself is the dominant feature the choice lies between limited surgery and resection. Retrograde dilatation may be of temporary value in the management of an awkward low stricture, especially if there are additional reasons for laparotomy. Intubation is probably best reserved for malignant strictures. Tubes tend to work loose from benign strictures after a time and reflux can be difficult to control (Holden and Wooler, 1971). Oesophagoplasty is a simple and effective method of relieving annular strictures in the lower oesophagus and avoids the need for resection in these cases. Good results were obtained in 6 out of 7 patients and postoperative reflux did not constitute a serious problem. The troublesome minority of long strictures (‘ascending oesophageal fibrosis’) will usually demand resection and some kind of oesophageal replacement. Allison (1970) has stated that direct oesophagogastric anastomosis merely reproduces the lesion that caused the original stricture, while Skinner and Belsey (1967) reported a mortality of 14.5 per cent in 106 patients SO treated. Twelve cases of oesophagogastrostomy are presented here without operative loss, though 3 of these needed postoperative medical treatment for reflux. Pyloroplasty was routinely performed. Raptis and Milne (1972) omitted this step in 41 oesophagogastrostomies performed for benign stricture and encountered anastomotic strictures in 10 of these. Eight of the 10 were found to have pyloric obstruction and were relieved by a subsequent pyloroplasty. Interposition of jejunum or colon must add to the morbidity and mortality of resection though it avoids the chance of any future reflux of acid peptic juice. Allison reported excellent results in 50 patients thus treated, with a mortality of only 4 per cent; jejunum was found to be more active than colon in conveying food.
Acknowledgements I am indebted to Mr Clement Grimshaw for allowing me to study the patients under his care at Peppard Hospital. I am also grateful to Professor M. H. Irving and Mr A. G . Cox for their help and encouragement in completing this paper. The illustrations were kindly provided by Mr Lionel Williams, Medical Photographer at the Royal Berkshire Hospital, Reading. References (1970) Peptic oesophagitis and oesophageal stricture. Lancet 2, 199-201. BARRETT N. R . (1962) Benign stricture in the lower oesophagus. J . Thorac. Cardiouasc. Surg. 43, 703-715. BENEDICT E. B. (1966) Peptic stenosis of the esophagus: a study of 233 patients treated with bouginage, surgery or both. Am. J . Dig. Dis.11, 761-770. BENEDICT E. B. and GILLESPIE J. E. O’N. (1954) Peptic stenosis of the esophagus. Surg. Cynecol. Obstet. 98, 494-502. BENNETT J. R . (1973) Symposium on gastro-oesophageal reflux: the physician’s problem. Gut 14, 246-249. ALLISON P. R.
453
R. C. N. Williamson EDITORIAL (1968)
The vulnerable oesophagus. Lancet
2, 267-268. and DAHLBACK 0. (1962) Instrumental perforation of the esophagus. Acta Otolaryngol. (Stockk.)54,279-286. GOLDSTEIN M. J. and SHERLOCK P. (1966) Silent esophageal perforation following esophagoscopy. Gasfrointest.Endosc. 13, 22-25. HARDIN w. J., HARDY J. D. and CONN J. H. (1967) Esophageal perforations. Surg. Gynecol. Obstet. 124, 325-331. HILL L. D., GELFAND M. and BAUERMEISTER D. (1970) Simplified management of reflux oesophagitis with stricture. Ann. Surg. 172, 638-651. HOLDEN M. P. and WOOLER G. H. (1971) MousseauBarbin tubes for benign strictures of the oesophagus. Thorax 26, 619-622. JOHKSON R. B. and LUKASH w. M. (1970) Dilatation of esophageal strictures. Mod. Treat. 7 , 1190-1203. KELLY A. B. (1936) Some oesophageal affections in young children. J . Laryngol. Otol. 51,78-99. ELNER A.
454
and GROVES L. K. (1970) Esophageal perforations. Ann. Thorac. Surg. 10,571-587. MARK J. B. D. and KNAUER c. M. (1969) Use of the onlay gastric patch in instrumental rupture of the esophagus. J . Thorac. Cardiovasc. Surg. 57, 883-886. PALMER E. D. (1968) The hiatus hernia-esophagitisesophageal stricture complex. Twenty-year prospective study. Am. J. Med. 44, 566-579. RAPTIS s. and MILNE D. M. (1972) A review of the management of 100 cases of benign stricture of the oesophagus. Thorax 27, 599-603. SKINNER D. B. and BELSEY R. H. R. (1967) Surgical management of esophageal reflux and hiatus hernia. Long term results with 1030 patients. J. Thorac. Cardiovasc. Surg. 53, 33-54. WYCHULIS A. R., FONTANA R. s. and PAYNE w. s. (1969) Instrumental perforations of the esophagus. Dis.Chest 55, 184-1 89. LOOP F. D.
The management of peptic oesophageal stricture R. C. N. W I L L I A M S O N * SUMMARY
Eighty-three adult pafients with peptic oesophageal stricture are reviewed with regard to diagnosis and treatment. Sliding hiatus hernia was present in 94 per cent of cases. The strictures were graded according to the degree of stenosis encountered at initial oesophagoscopy, but several othzr factors influenced the management of each case. Bouginage should always be attempted before proczeding to surgery. Thirty-eight patients were treated by bouginage alone, with 21 (55 per cent) good results and 3 deaths. Forty-five patients came to surgery, with 32 (71 per cent) good results and 3 deaths. Six patients sustained instrumental perforation of the oesophagus with I fatality. There was in addition 1 case of silent perforation following selfbouginage, which is believed 10 be unique. Hiatal herniorrhaphy combined with simple bouginage is recommended where reflux oesophagitis dominates the clinical picture. Local plastic repair (oesophagoplasty) has proved a simple and effective method of treating low annular strictures. Oesophagogastrostomy was carried out in 12 patients without operative loss and with good results in 8 cases; the addition of a pyloroplasty is believed to lessen postoperative reflux.
INJURYto the oesophagus may cause submucosal scarring, leading to permanent stenosis. In this country the commonest injurious agent is gastro-oesophageal reflux, but treatment of the resultant peptic stricture remains controversial. It has been stated that ‘bouginage should be the method of choice in every case unless one is dealing with severe haemorrhage or perforation’ (Benedict, 1966). Most British surgeons, however, advocate operation for the majority of cases (Editorial, 1968), though this may vary from local plastic repair to extensive resection with oesophageal replacement. The present communication reviews a large series of patients with peptic oesophageal stricture, with particular regard to their diagnosis and treatment in a district thoracic surgical unit.
Patients The series comprises 83 patients with peptic stricture treated at Peppard Hospital, Oxfordshire, during the Table I: AETIOLOGICAL TYPES OF BENIGN OESOPHAGEAL STRICTURE Aetiological type: Peptic NO. Of Cases: 83 Anastomotic 8 Sideropenic 2 Pemphigoid 1 Irradiation 1 Corrosive 1 -
Totdl
448
96
Fig. 1. Barium swallow film showing an early annular stricture of the lower oesophagus. This caused surprisingly severe symptoms but was successfully treated by local plastic repair (oesophagopldsty).
10-year period 1962-71 under the care of one consultant thoracic surgeon. Table I shows that this was by far the commonest aetiological type of benign oesophageal stricture seen during this period. All the patients were followed up for a minimum of 2 years after treatment. The total includes 49 women and 34 men, aged from 14 to 89 years with a mean of 68.2 years. The maximal incidence occurred during the seventh and eighth decades, two-thirds of patients being over the age of 70. Not surprisingly this elderly population had a high incidence of cardiorespiratory disease, 1 I patients being severely incapacitated.
* Royal Berkshire Hospital, Reading. Present address: United Bristol Hospitals.
Peptic oesophageal stricture
a
b
Fig. 2. a , ‘Ascending oesophageal fibrosis’. Barium swallow film showing an elongated peptic stricture. b, The appearance on a barium swallow film following resection of the stricture with jejunal replacement. The patient obtained an excellent result.
Sliding hiatus hernia was demonstrated in 78 of the 83 cases (94 per cent), while the remaining 5 patients all had clinical and endoscopic features of reflux oesophagitis. Three patients with no antecedent gastro-oesophageal symptoms developed an acute postoperative stricture, in each case following colonic surgery. Diagnosis Clinical Table II shows that dysphagia was the only constant symptom. It varied greatly in severity and duration and 7 patients presented with complete oesophageal obstruction (aphagia). It is notable that one-third of the patients had no previous history of reflux oesophagitis. Radiological Apart from 2 patients with aphagia who underwent emergency oesophagoscopy, all the strictures were first demonstrated by barium swallow examination. Figs. I and 2 illustrate the two classic types, namely annular stricture caused by a localized ring of fibrous tissue, and ascending oesophageal fibrosis (Kelly, 1936) where scarring extends for several centimetres with muscular hypertrophy, peri-oesophagitis and rnediastinal adenopathy. In 12 strictures subsequently shown to be benign there was radiological suspicion of
Fig. 3. Barium swallow film magnified to show a rounded filling defect in the lower oesophagus. This was reported as a leiomyoma but at thoracotomy the ‘tumour’ proved to be a peanut, swollen and impacted above an annular stricture. Oesophagoplasty was successfully performed. Table 11: SYMPTOMS OF PEPTIC STRICTURE No. of Symptom cases ;4 Dysphagia 83 100 Reflux oesophagitis (heartburn, 55 66 acid brash etc.) Regurgitation of solids 42 51 Weight loss (>,6,3 kg) 19 23 Haematemesis 8 9 Inhalation pneumonitis 5 6 Hiccuo 5 6
carcinoma; 5 of these also appeared irregular at endoscopy so that the possibility of carcinoma became a major factor in advising surgery. Fig. 3 demonstrates another pitfall in radiological diagnosis. Endoscopic In nearly every case the presence of a stricture was confirmed at oesophagoscopy and definitive treatment by bouginage was attempted at the same time. Some degree of oesophagitis was observed in 45 patients (54 per cent). Fig. 4 shows that the large majority of peptic strictures were confined to the lower third of the oesophagus, as expected. The strictures were graded according to the degree of stenosis encountered at the initial oesophagoscopy as follows: Grade I : Slight stenosis easily negotiated. Grade I I : Moderate stricture dilated without difficulty. Grade 111: Tight strictureonly dilated withdifficulty. Grade I V : Impassable stricture. 449
R. C. N. Williamson Distance from incisor teeth (cm)
Peptic strictures
Non-peptic strictures
ss
10
. .. .......... .... ..... ............ ...... ...... .... .......
Pg
R
tc A AAA
A AA A S=Sideropenic Pg= Pemphigoid R= Irradiation C=Corrosive A= Anastomotic
Fig. 4. Diagram to show the level of benign oesophageal strictures at endoscopy. Each peptic stricture is marked by a black circle. Strictures of non-peptic origin are included for comparison. The highest peptic stricture was found a t 18 cm from the incisor teeth; it proved amenable to bouginage.
Histologica I In 53 patients carcinoma was excluded as far as possible by the examination of biopsy or operative specimens, and in another 13 patients the oesophagus was directly inspected at thoracotomy. In the remaining 17 cases the diagnosis rested upon the continued absence of suspicious clinical, radiological or endoscopic features at follow-up. There was no incidence of a benign peptic stricture undergoing malignant change.
Methods of treatment The primary objective of treatment was to enable every patient to swallow sufficiently well to maintain his weight and nutritional status, albeit with some modification of diet. Associated oesophagitis was treated by a standard medical regime (Bennett, 1973). Dilatation of strictures was carried out using Neoplex or metal Souttar bougies, usually under general anaesthesia and always under direct endoscopic control, though 1 I patients whose stricture demanded repeated dilatation at frequent intervals were taught to pass their own bougies under a local anaesthetic. Fibre-optic oesophagoscopy was not available during the period of this study. The following surgical methods were employed. 450
Repair of hiatus hernia entailed exposure of the oesophagus within the chest, usually via a left thoracolaparotomy. Reduction of the hernia was achieved as far as possible and the hiatus was repaired by anterior apposition of the crura; routine HeinekeMikulicz pyloroplasty was carried out additionally. Operative dilatation of awkward low strictures was performed via a gastrotomy to allow digital dilatation or the retrograde passage of bougies. In a few such cases a Celestin tube was passed from above, right through the stricture and out, to ensure that adequate dilatation had been achieved. In certain selected cases the tube was left in situ. Oesophagoplasty or local plastic repair was reserved for low annular strictures which were exposed within the chest, incised longitudinally through all coats and closed transversely with one layer of interrupted catgut sutures, in a manner akin to Heineke pyloroplasty. Resection of the stricture was generally performed through a left thoracolaparotomy with direct oesophagogastric anastomosis and the addition of a pyloroplasty. Oblique transection of the stomach allowed the creation of a gastric tube for anastomosis with the transected oesophagus. In 3 cases a segment ofjejunum or right colon was brought up to bridge the defect following an extensive resection.
Peptic oesophageal stricture The long term result of treatment in each case was assessed as follows : Good: Manages virtually normal diet; minimal dysphagia. Moderate: Careful diet needed ; occasional regurgitation of food. Poor: Liquid diet; frequent dysphagia and regurgitation.
Where the result remained only ‘moderate’ after repeated bouginage surgery was advised unless contraindicated by extreme age o r infirmity, and these patients comprise the ‘bouginage ineffective’ group listed in Table IV. Forty-five patients were treated surgically (54 per cent) and all but 5 of these had undergone a previous trial of bouginage. Three types of surgical procedure were performed :
Results of treatment Of 55 patients presenting with symptoms of reflux, 24 continued to require medical treatment for this after either bouginage o r surgery. However, only 3 patients in all had symptoms of reflux aggravated by local treatment of the stricture. Table III shows that whatever the grading of the stricture at initial endoscopy roughly equal numbers of patients were treated by surgery and bouginage. Though grade I strictures responded readily to simple bouginage, surgery was often required in this group for severe associated reflux oesophagitis. The largest group of strictures belonged to grade 11; of these, half were treated surgically and half by bouginage, which usually became progressively easier. Grade I l l and IV strictures were very unpredictable. Five patients reported some improvement in swallowing despite unsuccessful bouginage, 2 of them requiring no further treatment. Five strictures, initially impassable, subsequently yielded t o bouginage, and the long term result in these difficult cases was sometimes surprisingly good. Nonetheless difficulties encountered during bouginage (including previous endoscopic perforation in 3 cases) were a common indication for surgery (Table I V ) if dysphagia subsequently recurred. The results of treatment are shown in Table V. Thirty-eight patients (46 per cent) were managed by bouginage alone, 17 requiring only one dilatation carried out at the time of diagnostic oesophagoscopy.
I . Repair of hiatus hernin A relatively young group of patients with marked oesophagitis was submitted to anti-reflux surgery with bouginage of the associated grade I o r 11 stricture. This combination was employed in 1 case of grade I11 stricture presenting in a girl of 14, but the result was poor and she required subsequent resection with oesophagogastrostomy. Table V shows that the results of this combined procedure were otherwise very Table 111: GRADING OF PEPTIC STRICTURES Grade
Bouainane
Surgery
II 44
5 22
9 11 8
4 6 1
6 22 5
83
38
I I1 111 1v Unclassified *
Total
Table IV: INDICATIONS FOR SURGERY Indication: Bouginage difficult N o. of cases: 14 Bouginage ineffective 13 12 Reflux oesophagitis Carcinoma suspected 5 ‘Leiomyoma’ (Fig. 3 ) 1 -
Total 45
Result Good
Moderate
Poor
Death
38 45 12 6 9 7 12 3
21 32
14 9 1 0 4 1 3
0 1 1
3 3 0 0 2 0 0
83
53 (64%)
10 5 3 6 8 2
0 23 (28%)
F F F F F F
1 0 0
1 0 1 ( 1%)
1 6 (7%)
one type of operative procedure was carried out and the final surgical result is recorded.
Table VI: CASES OF INSTRUMENTAL PERFORATION OF THE OESOPHAGUS Delay in onset of symptoms Level of Difficulty Age Sex at endoscopy (h) perforation Treatment 70 77 88 37 78 79
7
45
No. of cases
Total
* In 4 cases more than
5
* Stricture not seen at endoscopy or bouginage not attempted.
Table V: RESULTS OF TREATMENT Procedure Bouginage Surgery* Repair of hiatus hernia Retrograde dilatation Oesophageal intubation Oesophagoplasty Oesophagogastrostomy Jejunum/colon replacement
Treatment
N o. of cases
Postcricoid Postcricoid Nil Nil Grade 111 stricture Grade 111 stricture
24 24 24 24 4 2
Cervical Cervical Thoracic Thoracic Thoracic Thoracic
Conservative Conservative Conservative Conservative Operative Operative
Complications Nil Nil Empyema thordcis Peri-oesophageal abscess Nil Death
451
R. C. N. Williamson satisfactory within the period of follow-up, even though extensive peri-oesophagitis often prevented complete reduction of the hernia. The stricture was treated by prior endoscopic bouginage or, i n 2 cases, by retrograde digital dilatation at the time of operation. In 3 cases bouginage was required for a short time postoperatively as well. 2. Limited surgery to the stricture Operative dilatation was performed in 6 cases, including 1 man who underwent retrograde bouginage no less than three times in 4 months before obtaining a satisfactory and lasting result. The 1 patient with a ‘poor’ result had sustained a previous endoscopic perforation and was deemed too frail for any further procedure. Oesophageal intubation was reserved for a group of very elderly and unfit patients in whom bouginage was inadequate but major surgery impossible. It was not a particularly satisfactory procedure, for 2 patients died and several of the survivors had trouble with the tube. In 3 cases it slipped up the oesophagus to cause renewed dysphagia and require removal. One old lady passed her tube in two pieces per rectum to the consternation of the nursing staff. Oesophagoplasty was carried out in 7 cases of localized annular stricture close to the cardia (Figs. 1, 3). The results were good apart from 1 patient with residual dysphagia due to disordered oesophageal motility. Two patients with anastomotic stricture were treated with equal success by this procedure.
3 . Resection Oesophagogastrostomy was carried out in 12 cases; the oldest patient was 72 and all enjoyed reasonable general health. Although several had initial postoperative problems 8 patients obtained a ‘good’ result. Reflux oesophagitis spoilt the result in 3 cases, aggravated in one by obesity and in another by aspirin. Four patients needed limited postoperative bouginage. The only ‘poor’ result was in a patient who bled from an anastomotic ulcer, requiring further resection with jejunal interposition. This operation was performed in one other patient (Fig. 2) and neither case of jejunal replacement has required any further treatment. The sole case of colonic interposition died following an anastomotic leak on the tenth day, this being the only fatality among the 14 cases of resection. Complications Three patients died following endoscopic bouginage, a mortality of 7.9 per cent. The causes of death were pulmonary embolism, chronic chest infection and instrumental perforation. Two patients sustained a limited gastro-intestinal haemorrhage following instrumentation. There were 6 cases of instrumental perforation of the oesophagus during 270 endoscopies performed for benign stricture, an incidence of 2.2 per cent; the details are shown in Table VI. Strangely, 5 of the 6 patients had undergone previous uneventful 452
bouginage. In 2 cases dilatation was completed without suspicion of trauma, in 2 difficulty was encountered at the postcricoid level and in 2 the stricture itself was difficult to dilate. These last 2 both presented within a few hours of the accident and were treated surgically, though in I case operation was delayed for 48 hours and the patient died. The other 4 were treated conservatively with intravenous fluids and antibiotics in elixir form. Two of these developed a local abscess but made a complete recovery following drainage; the 2 patients with postcricoid perforation had minimal upset. The following case report illustrates one remarkable complication of self-bouginage : A 65-year-old man presented with dysphagia, epigastric pain and regurgitation of solids. Gross kyphosis was noted. Barium swallow examination showed a large hiatus hernia and an irregular stricture at the cardia with rigid margins suspicious of carcinoma. Endoscopy revealed an inflamed fibrous stricture at 35 cm which could not be dilated; biopsies showed no carcinoma. Ten days later the stricture was successfully dilated. The patient survived a major pulmonary embolus thereafter and his swallowing was much improved. In the following year the stricture was dilated three times and as it remained tight he was taught self-bouginage; this he practised on three occasions, the last time passing a no. 43 FG Neoplex bougie without difficulty. After this he began to develop abdominal pain and was readmitted with a small haematemesis 16 days after self-bouginage. Findings of peritonitis led to laparotomy and the discovery of a large left subhepatic abscess with a second smaller abscess within the liver. These were drained and the patient made an uneventful recovery. Two subsequent attempts a t endoscopic bouginage have failed but his swallowing is reasonably satisfactory.
Postoperatively, 3 patients died, a mortality of 6.7 per cent. Two deaths (from haematemesis and aspiration pneumonia respectively) followed insertion of a Celestin tube. The other fatality followed rupture of an interposed segment of colon. There was 1 case of postoperative chylothorax requiring thoracic duct ligation and one of strangulated diaphragmatic incisional hernia.
Discussion Reflux oesophagitis, the commonest disease affecting the gullet, may eventually lead to peptic stricture in 2-3 per cent of cases (Johnson and Lukash, 1970). Under certain circumstances (Allison, 1970), and especially after any major abdominal operation (Benedict and Gillespie, 1954), this process may develop rapidly; in the present series 3 patients developed an acute peptic stricture, in each case following colonic surgery. Benedict (1966) advocated bouginage for nearly every peptic stricture, reporting good results in 70 per cent of 133 patients. Many of these with unrepaired hiatal hernia became entirely symptom-free following bouginage. Barrett (1962) objected to simple bouginage on the grounds that it broke down a natural barrier against the upwards spread of oesophagitis, but in the present series reflux was only aggravated in 3 patients undergoing either bouginage or limited surgery to the stricture. Nevertheless, few surgeons would agree with Benedict that haemorrhage and perforation comprise the only indications for surgery. Only 55 per cent of
Peptic oesophageal stricture our patients obtained a good result from bouginage alone, and its failure to procure lasting relief from dysphagia was a common indication for operation. Likewise, difficulty encountered during previous bouginage and the possibility of carcinoma will usually demand surgical intervention. There were 6 cases of instrumental perforation during endoscopy, an incidence of 2.2 per cent, with 1 fatality. Raptis and Milne (1972) reported 4 such cases in their series of 100 benign strictures, again with 1 fatality. The overall incidence of perforation during all upper gastro-intestinal endoscopies is 0.5-0.7 per cent (Elner and Dahlback, 1962; Mark and Knauer, 1969), occurring most frequently in the cervical oesophagus in elderly women (Wychulis et al., 1969). This accident, of which 2 examples are reported here, can usually be managed conservatively (Hardin et al., 1967). Dilatation for benign or malignant stricture is the second commonest cause of iatrogenic perforation (Loop and Groves, 1970). The evidence suggests that patients who present with features of ruptured thoracic oesophagus within a few hours of endoscopy require immediate surgical exploration. The case of silent perforation during self-bouginage is believed to be without precedent, although Goldstein and Sherlock (1966) reported 1 case of unsuspected perforation following endoscopic bouginage carried out under a local anaesthetic. Endoscopic bouginage is therefore not without risk and in the present series carried a mortality (7.9 per cent) akin to that of operation (6.7 per cent). Furthermore, the 3 postoperative deaths occurred in patients in whom surgery had been necessitated by failure of bouginage. Taken in isolation, the tightness of the stricture was found to be a poor guide in determining the need for surgery. The natural history of peptic stricture appears to be unpredictable and the degree of reflux oesophagitis varies widely between cases. Age and infirmity often preclude the adoption of more vigorous methods of treatment, though it may be a matter of judgement whether multiple bouginage or limited surgery carries a greater risk in such cases. Twenty-one per cent of all cases required no other treatment than a single dilatation of the stricture carried out at the time of diagnostic oesophagoscopy. The policy was therefore to attempt bouginage in nearly every case, but if this should fail, to proceed to surgery, the choice of operation depending upon the degree of oesophagitis and the extent of the stricture. Where symptoms of reflux overshadow those of stricture lasting relief from dysphagia can usually be achieved by combining hiatal repair with simple bouginage (Hill et al., 1970). Palmer (1968) reported 7 patients treated in this way who subsequently required bouginage as often as those whose hernia had been left unrepaired, but in the present series this was not the case: of I 1 patients with early or moderate grades of stricture submitted to hiatal herniorrhaphy, only 3 needed any postoperative bouginage at all and that only for a brief period. This was in spite of the fact that gross peri-oesophagitis often prevented complete reduction of the hernia. 33
Where the stricture itself is the dominant feature the choice lies between limited surgery and resection. Retrograde dilatation may be of temporary value in the management of an awkward low stricture, especially if there are additional reasons for laparotomy. Intubation is probably best reserved for malignant strictures. Tubes tend to work loose from benign strictures after a time and reflux can be difficult to control (Holden and Wooler, 1971). Oesophagoplasty is a simple and effective method of relieving annular strictures in the lower oesophagus and avoids the need for resection in these cases. Good results were obtained in 6 out of 7 patients and postoperative reflux did not constitute a serious problem. The troublesome minority of long strictures (‘ascending oesophageal fibrosis’) will usually demand resection and some kind of oesophageal replacement. Allison (1970) has stated that direct oesophagogastric anastomosis merely reproduces the lesion that caused the original stricture, while Skinner and Belsey (1967) reported a mortality of 14.5 per cent in 106 patients SO treated. Twelve cases of oesophagogastrostomy are presented here without operative loss, though 3 of these needed postoperative medical treatment for reflux. Pyloroplasty was routinely performed. Raptis and Milne (1972) omitted this step in 41 oesophagogastrostomies performed for benign stricture and encountered anastomotic strictures in 10 of these. Eight of the 10 were found to have pyloric obstruction and were relieved by a subsequent pyloroplasty. Interposition of jejunum or colon must add to the morbidity and mortality of resection though it avoids the chance of any future reflux of acid peptic juice. Allison reported excellent results in 50 patients thus treated, with a mortality of only 4 per cent; jejunum was found to be more active than colon in conveying food.
Acknowledgements I am indebted to Mr Clement Grimshaw for allowing me to study the patients under his care at Peppard Hospital. I am also grateful to Professor M. H. Irving and Mr A. G . Cox for their help and encouragement in completing this paper. The illustrations were kindly provided by Mr Lionel Williams, Medical Photographer at the Royal Berkshire Hospital, Reading. References (1970) Peptic oesophagitis and oesophageal stricture. Lancet 2, 199-201. BARRETT N. R . (1962) Benign stricture in the lower oesophagus. J . Thorac. Cardiouasc. Surg. 43, 703-715. BENEDICT E. B. (1966) Peptic stenosis of the esophagus: a study of 233 patients treated with bouginage, surgery or both. Am. J . Dig. Dis.11, 761-770. BENEDICT E. B. and GILLESPIE J. E. O’N. (1954) Peptic stenosis of the esophagus. Surg. Cynecol. Obstet. 98, 494-502. BENNETT J. R . (1973) Symposium on gastro-oesophageal reflux: the physician’s problem. Gut 14, 246-249. ALLISON P. R.
453
R. C. N. Williamson EDITORIAL (1968)
The vulnerable oesophagus. Lancet
2, 267-268. and DAHLBACK 0. (1962) Instrumental perforation of the esophagus. Acta Otolaryngol. (Stockk.)54,279-286. GOLDSTEIN M. J. and SHERLOCK P. (1966) Silent esophageal perforation following esophagoscopy. Gasfrointest.Endosc. 13, 22-25. HARDIN w. J., HARDY J. D. and CONN J. H. (1967) Esophageal perforations. Surg. Gynecol. Obstet. 124, 325-331. HILL L. D., GELFAND M. and BAUERMEISTER D. (1970) Simplified management of reflux oesophagitis with stricture. Ann. Surg. 172, 638-651. HOLDEN M. P. and WOOLER G. H. (1971) MousseauBarbin tubes for benign strictures of the oesophagus. Thorax 26, 619-622. JOHKSON R. B. and LUKASH w. M. (1970) Dilatation of esophageal strictures. Mod. Treat. 7 , 1190-1203. KELLY A. B. (1936) Some oesophageal affections in young children. J . Laryngol. Otol. 51,78-99. ELNER A.
454
and GROVES L. K. (1970) Esophageal perforations. Ann. Thorac. Surg. 10,571-587. MARK J. B. D. and KNAUER c. M. (1969) Use of the onlay gastric patch in instrumental rupture of the esophagus. J . Thorac. Cardiovasc. Surg. 57, 883-886. PALMER E. D. (1968) The hiatus hernia-esophagitisesophageal stricture complex. Twenty-year prospective study. Am. J. Med. 44, 566-579. RAPTIS s. and MILNE D. M. (1972) A review of the management of 100 cases of benign stricture of the oesophagus. Thorax 27, 599-603. SKINNER D. B. and BELSEY R. H. R. (1967) Surgical management of esophageal reflux and hiatus hernia. Long term results with 1030 patients. J. Thorac. Cardiovasc. Surg. 53, 33-54. WYCHULIS A. R., FONTANA R. s. and PAYNE w. s. (1969) Instrumental perforations of the esophagus. Dis.Chest 55, 184-1 89. LOOP F. D.
