Anastomotic
Stricture
By Soottiporn
Following
Chittmittrapap,
Lewis Spitz, Edward London,
l Anastomotic strictures developed in 74 (37.2%) of 199 patients undergoing primary or delayed primary repair of esophageal atresia with or without tracheoesophageal fistula. Significant predisposing factors included the use of braided silk sutures (relative risk 1.72 and 1.49, compared with polyglycolic acid and polypropylene sutures), the presence of gastroesophageal reflux and leakage of the anastomosis (relative risk 2.29 and 2.04. respectively). Tracheomalacia, personnel factors, and recurrent fistula did not affect the rate of stricture formation. Seventy-one patients responded to dilatation alone, whereas three required stricture resection or esophageal substitution. All three patients requiring surgical intervention and 14 requiring five or more dilatations developed symptoms within the first 6 months after esophageal anastomosis. Antireflux surgery was carried out in 19 (26.7%) of the 74 patients. 0 1990 by W.B. Saunders Company. INDEX WORDS: Esophageal fistula; esophageal stricture.
atresia;
Repair of Esophageal
Atresia
M. Kiely, and R.J. Brereton
England
and analyzed for possible risk factors in 194 patients who underwent primary repair, and five who underwent delayed primary repair of esophageal atresia. Details of anastomotic strictures are presented here, although details concerning Ieakage and recurrent fistula wiII be subjects of separate publications. Patients who experienced swallowing problems (dysphagia, vomiting, poor or slow feeding) or recurrent episodes of respiratory infection or foreign body (including food bolus) obstruction, in whom a stricture that required dilatation was demonstrated at the time of endoscopic examination, were included in the “stricture” group. Patients admitted to the hospital with the previously mentioned symptoms, predominantly those with foreign body or focd bolus obstruction, who at endoscopy were shown to have normal esophageal lumen and did not require dilatation, were classified as a “questionable” group. Follow-up period ranged from 4 months to 7 years. Chi-squared and Student’s t test were used for the statistical analysis. RESULTS
tracheoesophageal
Of the 199 patients, 15 were excluded from the analysis; six due to early mortality, six in whom the NASTOMOTIC STRICTURES, leakage at the anastomosis disrupted requiring cervical esophagosanastomotic site, and recurrent tracheoesophtomy and subsequent esophageal replacement, two ageal fistula are common complications following repatients were lost to follow-up, and one infant for pair of esophageal atresia. Depending upon the criteria whom the follow-up period following delayed primary for the definition of stricture formation, the incidence repair was too short (t2 months) for evaluation. One of this complication varies widely from 8%’ to 49%~~~~ hundred two patients demonstrated no evidence of Many predisposing factors have been implicated, instricture formation. There were 74 patients in the cluding the use of silk sutures for the anastomosis,5 stricture group and eight in the questionable group. Haight and two-layer anastomosis,6’7 end-to-side The effect of suture materials used for the anastomoanastomosis,’ anastomotic tension,’ and gastroesophsis on the subsequent development of anastomotic ageal reflux. 4~‘oStatistical analysis of the various risk stricture (without regression of gastroesophageal refactors has been difficult to achieve because of the flux factor) is shown in Table 1. Ethibond (Ethicon small number of patients attending one center during a Ltd, Edinburgh, Scotland) was used in only 12 cases relatively short defined time scale. At the Hospital for and strictures occurred in three; because the numbers Sick Children, London, 30 to 35 newly diagnosed cases were small it was not further evaluated. There was a of esophageal atresia are admitted annually. In the significant increase in stricture rate in cases in which 8-year period (1980 to 1987) 253 patients with esophsilk sutures were used (P c .05) both with or without ageal atresia were available for analysis and they form regression of the reflux factor. The relative risk for silk the core material for this paper.
A
MATERIALS
AND
Table 1. Suture Materials Used for Repair and Incidence of Stricture
METHODS
The medical records of 253 infants admitted with the diagnosis of esophageal atresia and/or tracheoesophageal fistula were retrospectively reviewed. All anastomotic complications including stricture, leakage, and recurrence of tracheoesophageal fistula were recorded
NO
Silk
Not
Stricture
Stricture
Ouestionable
Evaluated
Total
15
12
1
2
30
22
46
3
7
78
33
36
3
5
77
3
7
1
1
12
1
1
0
0
2
8
15
199
Polyglycolic acid
From The Hospital for Sick Children, London. England. Date accepted: February 9.1989. Address reprint requests to Professor L. Spitz, MD, Institute of Child Health. 30 Guilford St, London WC1N I EH, England. 8 1990 by W.B. Saunders Company. 0022-3468/90/r505-0010$03.00/0 608
Polypropylene Ethibond No data of suture
74
Total
Journal
of Pediarric
Surpry,
102
Vol 25, No 5 (May), 1990: pp 508-511
509
STRICTURE AFTER ESOPHAGEAL ATRESIA REPAIR
Table 2. The Preaenco of Gastroeaophageal Stricture
Stricture
NO Strictwe
GER
39.
22t
No GER
30
80
5
0
Not evaluated
Table 4. Tracheomalacia
Reiiux (GEM and
NO Stricture
Not Qusationabb
4
Evaluated
Total
0
65
4
3
117
0
12
17
l
74
102
8
15
199
Stricture
Ouestbnsble
Tracheomalacia
23,
1st
1
No tracheomalacia
51
75
7
NOTE. Only 175 of 184 were evaluated for tracheomalacia. l
Total
and Stricture
Formation
16 required aortopaxy.
t 14 required aortopexy.
19 required Nissen’s fundoplication.
t 15 required Nissen’s fundoplication.
was I .72 and 1.49, compared with polyglycolic acid and polypropylene sutures, respectively. A total of 22 strictures were noted in the 78 cases in which polyglycolic acid was used, compared with 33 strictures in 77 cases with polypropylene sutures. But, with the regression of gastroesophageal reflux factor, no statistically significant difference between polyglycolic acid and polypropylene sutures could be demonstrated. Table 2 shows the effect of the presence and absence of gastroesophageal reflux (without elimination of silk suture effect). Gastroesophageal reflux significantly increased the stricture rate (PC .Ol) both with or without regression of the silk-suture factor. The relative risk of gastroesophageal reflux was 2.29. Leakage at the anastomotic site also significantly affected the stricture rate (P < .05); the relative risk was 2.04 (Table 3). The stricture rate was unaffected by tracheomalacia (Table 4) or by recurrent tracheoesophageal fistula (Table 5). There was no difference in the incidence of stricture relating to the status of the surgeon (Table 6). Dilatation of the stricture was the initial treatment of choice and was successful in 71 of the 74 patients. Twenty-one patients (28.4%) required only a single dilatation, 21 required two dilatations (28.4%), seven required three dilatations (9.5%), eight required four (10.8%), and 14 patients (18.9%) needed five or more dilatations. Prophylactic dilatations were not carried out as a routine. Two of the 74 patients had dilatations and then resection later on. Gastric interposition was performed in another child who had both stricture and recurrent fistula. The presenting symptoms and the number of dilatations required to resolve the problems are shown in Table 3. Anastomotic
Leakage
Correlated
NO Stricture
Stricture
Recent refinements of operative technique, improvement in preoperative and postoperative management, and anesthesia of the newborn infant have directly contributed to the improved survival rate for infants born with esophageal atresia. The standard approach to such lesions is directed to primary repair in all cases,
Not Questionable
Evaluated
Total
20
7
0
7
34
No leak
54
95
8
3
160
0
0
0
5
5
74
102
8
15
199
Total
DISCUSSION
to Stricture
Leak
Not evaluated
Table 7. Swallowing difficulties and/or poor feeding were the most prominent symptoms (79.7%). Food bolus impaction above the stricture occurred in six children, all over the age of 1 year, and all responded to one to two dilatations. The age at onset of the first symptoms is also shown in Table 7. The earlier the onset of symptoms, the more dilatations were required. All patients requiring five or more dilatations or surgical correction of the stricture developed their symptoms within the first 6 months postoperatively. The association between stricture and gastroesophageal reflux is shown in Table 8. The incidence of significant gastroesophageal reflux and the requirement for antireflux surgery increased in proportion to the number of dilatations required. Thirty-nine of 74 stricture patients had reflux but only 19 patients underwent Nissen’s fundoplication, 18 of whom presented with clinical features of stricture within the first 6 months postoperatively. Perforation of the esophagus occurred as a complication of dilatation in three patients. The perforation sealed on conservative treatment in one infant, but further dilatation was required for a recurrent stricture. Surgical exploration with repair of the perforation and fundoplication was carried out successfully in one infant. The third patient underwent lateral esophagostomy but died later, following cardiac surgery. A recurrent tracheoesophageal fistula was unmasked after dilatation in one patient.
Table 5. Recurrent
Tracheoesophageal
and Stricture
ITE) Fistula
Formation NO
Stricture
Stricture
Questionable
Recurrent TE fistula
5
10
0
No recurrent fistula
69
92
8
510
CHIT-TMITTAAPAP ET AL
Table 6. Personnel
Factor and Stricture
Table 6. Gastroesophageal
No Stricture
Stricture
Questionable
Operated by consultants
42
60
4
Operated by trainees
30
42
4
No. of Dilations Required
NOTE. Thare is no record on the personnel factor in two of 184 cases.
wherever feasible. The current survival rate in specialized centers is 80% to 90%.iV5 In this study of 253 patients treated over an g-year period, the overall survival rate was 85.8%. Excluding the infants with other severe congenital anomalies incompatible with survival, the survival rate was 89.3%. Anastomotic complications are important causes of postoperative morbidity. The nature of the suture materials used in the anastomosis has been claimed to play an important role in the development of a subsequent stricture. Compared with other suture materials, braided silk was associated with a greater incidence of stricture formation.5.7 Touloukian’ relied on fine (5/O) silk sutures and was able to report a low stricture rate of 6%. In an experimental study on esophageal anastomosis, Kullendorff et al” concluded that there was no correlation between the degree of stricture formation and the remnant of suture material within the anastomosis. The stricture rate was unrelated to the use of absorbable or nonabsorbable suture material. The size of the suture used in the anastomosis could not be correlated with stricture formation. Kullendorff et al defined stricture as the difference in luminal circumference at and below the anastomosis. This does not correlate with clinical stricture, and in esophageal atresia there is always a difference in circumference at the anastomosis. In this series, the size of sutures used for anastomosis were 5/O and 6/O, which were similar in each group of suture materials so this was not evaluated. This study has clearly shown an increased incidence of stricture formation associated with the use of braided silk, whereas polyglycolic acid and polypropylene sutures produced similar stricture rates. Gastroesophageal reflux has been implicated in the pathogenesis of anastomotic stricture by causing pep Table 7. The Presenting
Reflux and Antireflux
the Stricture
Symptoms
Operation
No. of Cases
Presence of Reflux
No. Nissen Performed
1
21
4
3
2
21
12
6 2
3
7
6
4
8
4
1
14
11
7
3
2
0
74
39
19.
>5 Resection/ substitution Total
*Onset less than 6 months, 18 cases: onset greater than 6 months, one case.
tic esophagitis in a vulnerable area. If the reflux is not aggressively treated, dilatation alone is rarely successfu1.4*‘o*12 Our study showed a statistically significant relationship between reflux and stricture formation, After antireflux surgery, the majority of strictures resolved with minimal interference. Similarly, leakage at the anastomotic site was significantly related to stricture formation, although this contrasts with other series. ’ 3 The development of recurrent tracheoesophageal fistula did not affect the caliber of the anastomosis, although five patients developed concurrent strictures and recurrent fistulae. The most common symptom of stricture was swallowing difficulty. In the absence of a definite stricture, impaired esophageal motility was the most likely cause. Strode1 et alI4 and Louhimo and Lindahl” routinely performed prophylactic esophageal dilatation postoperatively, followed by regular sounding of the anastomosis without regard to the infant’s symptoms. In these cases it is impossible to assess the incidence of stricture, and the role of prophylactic dilatation in reducing the incidence of stricture is unproven. The asymptomatic patients are at risk of developing complications during dilatation. In this series, dilatation was used only in symptomatic patients and was successful in 95% of the cases,
of the First Attack
Related to Type of Treatment Onset (months)
Symptomscr Problems D-6
z-6
4
10
11
3
2
15
6
7
0
0
5
2
4
8
0
0
5
3
z-5
10
4
0
14
0
Resection/substitution
3
0
0
3
0
59
9
6
52
22
No. of DilationsRequired
Swallowing
1
15
2
2
16
3
Total
Respiration
in
Group
FoodObstruction
511
STRICTURE AFTER ESOPHAGEAL ATRESIA REPAIR
REFERENCES 1. Touloukian RJ: Long term results following repair of esophageal atresia by end-to-side anastomosis and ligation of the tracheoesophageal fistula. J Pediatr Surg 16:983-988,198l 2. Lundertse-Verloop K, Tibboel D, Hazebrock FWJ, et al: Postoperative morbidity in patients with esophageal atresia. Pediatr Surg Int 2:2-S, 1987 3. Holder TM, Ashcraft KW, Sharp RJ, et al: Care of infants with esophageal atresia, tracheoesophageal fistula and associated anomalies. J Thorac Cardiovasc Surg 94:828-8351987 4. Holder TM: Esophageal atresia and tracheoesophageal fistula, in Ashcraft KW, Holder TM (eds): Pediatric Esophageal Surgery. Philadelphia, PA, Grune & Stratton, 1987, pp 29-52 5. Spitz L, Kiely EM, Brereton RJ: Esophageal atresia: Five years experience with 148 cases. J Pediatr Surg 22:103-108, 1987 6. Hicks LM, Manfield PB: Esophageal atresia and tracheoesophageal fistula. Review of thirteen years’ experience. J Thorac CardiovascSurg 81:358-363, 1981 7. Holder TM, Cloud DT, Lewis JE Jr, et al: Esophageal atresia and tracheoesophageal fistula: A survey of its members by the surgical section of the American Academy of Pediatrics. Pediatrics 34542-549, 1964 8. Ein S, Ashcraft KW. In discussion: Touloukian RJ: Long term
results following repair of esophageal atresia by end-to-end anastomosis and ligation of the tracheoesophageal fistula. J Pediatr Surg 16:983-988, 1981 9. Jolly SG, Johnson DG, Roberts CC, et al: Patterns of gastroesophageal reflux in children following repair oP esophageal atresia and distal tracheoesophageal fistula. J Pediatr Surg 15:857-862, 1980 10. Aprigliano F: Esophageal stenosis in children. Ann Otol 89:391-395,198O 11. Kullendorff CM, Okmian L, Jonsson N: Technical considerations of experimental esophageal anastomosis. J Pediatr Surg 16:979-982, 1981 12. Lindahl H, Louhimo I: Livaditis myotomy in long-gap esophageal atresia. J Pediatr Surg 22:109-l 12, 1987 13. Holder TM, Ashcraft KW: Esophagus, in Welch KJ (ed): Complications of Pediatric Surgery. Prevention and Management. Philadelphia, PA, Saunders, 1982, pp 199-207 14. Strode1 WE, Coran AG, Kirsh MM, et al: Esophageal atresia: A 41-year experience. Arch Surg 114:523-527.1979 15. Louhimo I, Lindahl H: Esophageal atresia: Primary result of 500 consecutively treated patients. J Pediatr Surg 18:217-229, 1983