Ann Thorac Surg 2002;73:267-272
© 2002 The Society of Thoracic Surgeons
Original article: general thoracic
Esophageal atresia: historical evolution of management and results in 371 patients
Jacqueline A. Deurloo, MDa,
Seine Ekkelkamp, MDa,
Mak Schoorl, MDa,
Hugo A. Heij, MD, PhDa,
Daniel C. Aronson, MD, PhD*a
a Pediatric Surgical Center of Amsterdam, Emma Children’s Hospital Academic Medical Center, and Vrije Universiteit Medical Center, Amsterdam, The Netherlands
Accepted for publication August 22, 2001.
* Address reprint requests to Dr Aronson, Pediatric Surgical Center of Amsterdam, Emma Children’s Hospital AMC, Academic Medical Center, PO Box 22660, 1100 DD, Amsterdam, The Netherlands
e-mail: d.c.aronson{at}amc.uva.nl
 |
Abstract
|
|---|
Background. It has been more than 50 years since the first successful surgical reconstruction of esophageal atresia was performed in The Netherlands. We reviewed the historical changes in management and treatment results of patients born with esophageal atresia.
Methods. We developed and analyzed a database of 371 consecutive patients treated for esophageal atresia in our center between 1947 and 2000.
Results. The mean birthweight decreased from 2,723 g (1947 to 1968) to 2,494 g (1994 to 2000), the mean gestational age decreased from 39 weeks (1947 to 1968) to 37 weeks (1994 to 2000). The number of patients with associated congenital malformations increased from 34% (1947 to 1968) to 66% (1994 to 2000). Most patients underwent primary repair of their atresia. Clinically significant tracheomalacia was present in 34 of 269 patients (13%). Gastroesophageal reflux was present in 90 of 277 patients (33%). Mortality decreased from 61% (1947 to 1968) to 11% (1994 to 2000).
Conclusions. The patients who are treated nowadays for esophageal atresia in a pediatric surgical center are born earlier, weigh less, and have more associated anomalies than those treated 50 years ago. Still, the mortality rate is much lower thanks to earlier diagnosis, better supportive care and improved surgical techniques. Therefore, further significant reduction will be difficult to achieve.
|
Introduction
|
|---|
More than 50 years ago, Eerland [1] performed the first successful surgical reconstruction of esophageal atresia in The Netherlands. In the following decades, the survival of children with esophageal atresia has improved drastically. Several factors may influence the result of treatment of a child with esophageal atresia. These factors include advances in surgical techniques, advances in preoperative and postoperative medical care, and changes in patient characteristics. The aim of this paper is to review the historical changes in the management of patients born with esophageal atresia, and the impact of these changes on mortality and postoperative results. To do so, we developed and analyzed a database of 371 consecutive patients treated for esophageal atresia in our center between 1947 and 2000.
|
Material and methods
|
|---|
Between September 1947 and March 2000, 371 patients with esophageal atresia or tracheoesophageal fistula or both were treated at the Pediatric Surgical Center of Amsterdam. The following data were collected retrospectively from hospital charts, operative reports, and office notes, and were entered in a database: demographics, type of atresia, polyhydramnios, associated congenital malformations, pneumonia, Waterston classification, type of operation, occurrence of pleural lesion, anastomotic tension (as reported by the surgeon in the operative report), postoperative complications and treatment, length of hospital stay, hypertrophic pyloric stenosis, symptoms of gastroesophageal reflux (GER), 24-hour pH measurement, esophagogastroscopy, esophageal biopsies, antireflux procedures, symptoms of tracheomalacia, bronchoscopy, aortopexy, dilatations of anastomotic stricture, recurrent pulmonary problems, growth and (if applicable) cause of death. For further analysis, the cause of death was categorized into five groups: surgical-technical complications (eg, anastomotic leak), nonsurgical complications (eg, pneumonia), associated congenital malformations, other cause of death, and unknown cause of death.
Analysis was performed using SPSS (Statistical Products and Service Solutions) 8.0.2 for Windows (SPSS Inc, Chicago, IL). Differences between groups were analyzed by means of the 
2-test and one-way analysis of variance (ANOVA) test. The level of significance was defined as p = 0.05. To analyze historic trends, the database was divided into five time periods with approximately equal numbers of patients (1947 to 1968, 1969 to 1977, 1978 to 1985, 1986 to 1993, and 1994 to 2000). The Waterston classification was used to distinguish three risk groups characterized by birthweight, preoperative pneumonia, and associated congenital malformations. The birthweight of 14 infants could not be retrieved, which made classification according to Waterston impossible in these cases.
|
Results
|
|---|
Patient characteristics
The study group consisted of 215 boys (58%) and 155 girls (42%). The sex of 1 patient could not be determined owing to ambiguous genitalia, and this patient died shortly after birth. The mean birthweight was 2552 g, range 650 to 4160 g; 32 patients (9%) had a birthweight less than 1,500 g. The mean gestational age was 38 weeks, range 26 to 43 weeks; 114 patients (31%) were born prematurely with a gestational age less than 37 weeks. Over the five time periods, the mean birthweight decreased from 2,723 g (1947 to 1968) to 2,494 g (1994 to 2000; p = 0.33) and the mean gestational age decreased from 39 weeks (1947 to 1968) to 37 weeks (1994 to 2000; p = 0.01).
Table 1
shows the distribution of the types of atresia according to the classification of Gross. Polyhydramnios was recorded in a total of 176 of 371 patients (47%): in 20 of 25 patients (80%) with type A atresia, in 2 of 2 patients (100%) with type B atresia, in 152 of 329 patients (46%) with type C atresia, in 1 of 2 patients (50%) with type D atresia, and in 1 of 13 patients (8%) with type E atresia.
Comorbidity
One hundred seventy-five patients (47%) had no other congenital malformations, the remaining 196 patients (53%) had one or more associated congenital malformations (Table 2).
The changes in prevalence of malformations in the different subgroups over the five time periods are given in Table 3.
The number of patients with associated congenital malformations has increased from 34% (24 of 71, 1947 to 1968) to 66% (51 of 71, 1994 to 2000; 2 for trend: p < 0.0001). For instance, the number of cardiac malformations has increased from 11% (8 of 71, 1947 to 1968) to 42% (32 of 77, 1994 to 2000; 2 for trend: p = 0.001). At least part of this increase can be explained by more sophisticated methods of evaluation.
View this table:
[in this window]
[in a new window]
|
Table 2. Specification of Congenital Malformations in 196 Patients Born With Esophageal Atresia and Associated Congenital Malformations
|
|
Pneumonia was present preoperatively in 59 patients (17%). The number of patients with preoperative pneumonia has decreased over the five time-periods from 15% (10 of 69 patients, 1947 to 1968) to 8% (6 of 76, 1994 to 2000; 2 for trend: p = 0.032).
Operations and complications
An operation was performed in 357 patients, 14 patients died without surgical intervention because of poor clinical condition or major additional congenital malformations. The operations performed according to the different types of atresia are listed in Table 4.
Four patients with type A atresia did not undergo surgical intervention because of major additional congenital malformations. These 4 patients all had trisomy 21 and cardiac malformations. In 2 patients, both operated upon in 1949, the distal esophageal segment could not be found during operation. They both died shortly after the operation. The other 19 patients underwent gastrostomy as initial surgical procedure. Six of these patients died without further surgical interventions. Thirteen patients underwent esophageal reconstruction at a later date. We prefer to use the native esophagus for this reconstruction whenever possible; this was achieved in 10 of 13 patients. Primary gastric interposition was performed in 1 patient (in 1977), colonic interposition was performed in 2 patients (in 1968 and 1971). A secondary gastric interposition was performed in 1 patient after failure of the primary reconstruction.
Both patients with type B atresia underwent gastrostomy as initial surgical procedure. Esophageal reconstruction was performed at a later date in these patients.
In 295 of 329 patients with type C atresia, reconstruction was possible during the initial operation. In 10 of these patients esophageal elongation was necessary (8 myotomies and 2 Ten Kate-procedures). In 13 patients a gastrostomy was performed as initial procedure. Seven of these patients died without further surgical interventions. In 3 of 13 patients delayed esophageal reconstruction was performed, and 2 of 13 patients died after a second operation during which only the fistula was closed as the clinical condition of the patient did not allow esophageal reconstruction (in 1966 and 1967). In 1 patient a gastrostomy was performed as initial surgical procedure, in combination with ligation of the fistula. A reconstruction was performed at a later date. In 9 of 329 patients the fistula was closed during the first operation without further reconstruction; the operation had to be ended because of the poor clinical condition of the patients or because the distal part of the esophagus could not be found (in 1949 and 1950). Five of these 9 patients died without further surgical interventions, 2 of 9 patients underwent gastrostomy for feeding but died shortly after this second operation. A colonic interposition was performed in the other 2 patients. Twelve patients with type C atresia died without any surgical intervention.
Both patients with type D atresia underwent primary reconstruction. In 1 of these patients there was a diagnostic delay because the nasogastric tube was positioned into the stomach through both fistulas.
In 12 of 13 patients with type E atresia the fistula was divided through a cervical approach. For unknown reasons, a gastrostomy was performed in 1 patient (in 1970). This patient died shortly after the operation.
In 38 of 317 evaluable cases (12%) the operative report mentioned anastomotic tension. A pleural lesion occurred in 125 of 317 evaluable patients (39%). Since approximately 1968, an extrapleural reconstruction was performed as standard procedure. A pleural lesion occurred in 92 of 247 patients (37%) operated on since 1968.
A total of 149 postoperative complications occurred in 131 of 357 patients (37%). An anastomotic leak was reported in 34 patients (9%), 1 of these patients required a second operation because of the anastomotic leak. An anastomotic leak developed more often in patients with reported anastomotic tension (9 of 38 versus 25 of 319, p < 0.005). The number of anastomotic leaks has decreased over the five time periods from 11% (1947 to 1968) to 4% (1996 to 2000; not significant). Pneumothorax was reported in 15 of 357 patients (4%), drainage was performed in 14 of these patients. A pneumothorax developed more often in patients with a pleural lesion (11 of 125 versus 4 of 232, p < 0.004). Postoperative pneumonia was reported in 23 of 357 patients (6%), which was treated with antibiotics in 17 of these patients. Since approximately 1968, an extrapleural approach was used instead of a transpleural approach. We found no significant change in the prevalence of a postoperative anastomotic leak or pneumothorax.
Recurrence of a tracheoesophageal fistula was reported in 5 patients (2%), all with type C atresia. Two patients developed a tracheoesophageal fistula after repair of type A atresia. The fistula was closed endoscopically with tissue adhesive in 1 patient, the other patients were treated operatively; none of them had a second recurrence of a tracheoesophageal fistula. There was no correlation between anastomotic tension and recurrent fistula or between pleural lesion and recurrent fistula.
Postoperative morbidity
A hypertrophic pyloric stenosis developed in 16 of 288 evaluable patients (6%). The normal distribution of the different types of atresia was seen in these 16 patients (14 patients had type C atresia, 2 patients had type A atresia).
A 24-hour pH measurement was performed in 120 of 269 surviving patients (45%), demonstrating GER in 48 of 120 patients (40%). GER was diagnosed in a total of 87 of 269 surviving patients (32%), based on clinical symptoms and 24-hour pH measurements. GER was more often diagnosed in patients with a type A and type B atresia (with a long gap) than in patients with the other types of atresia (9 of 14 patients with type A/B versus 78 of 255 patients with type C/D/E, p < 0.05). There was no correlation with anastomotic tension or a pleural lesion during operation.
Since 1986, patients are screened routinely for GER by means of pH measurements 3 months after the esophageal reconstruction. Table 5
shows the results of this routine on the number of patients diagnosed with GER, number of dilatations, and number of antireflux procedures.
View this table:
[in this window]
[in a new window]
|
Table 5. Results of Routine pH Measurements (Since 1986) on Number of Patients Diagnosed With GER, Number of Dilatations, and Number of Gastropexies
|
|
A Boerema anterior gastropexy is our standard antireflux procedure [2–4]. The aim of this operation is to bring a segment of esophagus intraabdominally, and to normalize the angle of His. This is done by mobilizing the esophagus transhiatally and fixating the lesser curvature of the stomach to the anterior abdominal wall. This procedure was performed in 61 of 269 surviving patients (23%). Failure of maximum conservative therapy for GER was the indication in 58 of 61 patients, an apparent life-threatening event in 2 of 61 patients, and a recurrent anastomotic stenosis in 1 of 61 patients. A second antireflux procedure was necessary in 7 patients (11%) for recurrence of GER. A second Boerema anterior gastropexy was performed in all patients, combined with a Nissen-fundoplication in 3 patients.
An anastomotic stenosis requiring more than 3 dilatations developed in 47 of 269 surviving patients (17%). GER was diagnosed in 22 of these patients (47%), 14 of whom (64%) underwent anterior gastropexy.
Growth was normal in 250 of 269 surviving patients (93%), with height and weight above the fifth percentile for age (P5). Height and weight were below the fifth percentile for age (P5) in 19 of 269 surviving patients (7%). Patients with GER more often had impaired growth (< P5; 13 of 87 versus 6 of 182; p < 0.005). Also patients with birthweight less than 1500 g more often had impaired growth (6 of 17 versus 13 of 252; p < 0.0005). No correlation between type of atresia and impaired growth could be found.
Clinically significant tracheomalacia was diagnosed in 34 of 269 surviving patients (13%), based on clinical symptoms and tracheobronchoscopy. An aortopexy was performed in 18 of 34 patients (53%), with good results in all patients. Thirteen of these 18 patients also underwent a gastropexy. Since 1986 gastropexy was always performed before aortopexy after clinical experience had shown that treatment of GER often eradicated the symptoms of tracheomalacia. Before 1986 more patients diagnosed with tracheomalacia underwent aortopexy but these numbers are too small to make a statistically significant difference (9 of 16 versus 9 of 21; p = 0.42). Patients with GER more often had tracheomalacia (25 of 87 versus 9 of 182; p < 0.0001). Recurrent respiratory problems occurred in 87 of 269 surviving patients (32%). Patients with GER more often had recurrent respiratory problems (45 of 87 versus 42 of 182; p < 0.0001).
Mortality
Table 6
shows the mortality in the three risk groups according to Waterston over the five periods. Overall mortality has decreased from 61% to 11% (2 for trend: p < 0.0001). In the most recent time period practically all patients from Waterston group A and group B survived, whereas in the first time period mortality in these subgroups was 50% and 73%, respectively.
The causes of mortality over the five time periods are shown in Table 7.
Surgical-technical complications were the most frequent cause of death in the first time period and did not cause any mortality in the last time period. A striking example of a surgical complication is the finding that in the early years the surgeon sometimes could not detect the distal esophagus. Associated congenital malformations were the most important cause of death in the last time period, causing 78% of mortality, although the absolute figure hardly changed: 6 of 71 patients (8%) from 1947 to 1968 and 7 of 77 (9%) patients from 1994 to 2000.
|
Comment
|
|---|
The mortality of patients born with esophageal atresia has decreased from 61% to 11% since 1947. In the most recent years 78% of mortality is caused by associated congenital anomalies. This is in sharp contrast to the earlier years when mortality was largely due to surgical-technical and nonsurgical causes. The mortality of esophageal atresia seems to be mostly determined by its comorbidity and a further reduction in mortality will depend on the treatment of the associated congenital anomalies.
The reduction in mortality has been achieved even though patients who are treated for esophageal atresia in a pediatric surgical center seem different from those treated 50 years ago. Today these patients are born earlier, weigh less, and have more associated anomalies. Still, fewer patients have pneumonia at the time of operation and the treatment results are better: there are fewer surgical complications and the survival rate has markedly improved over the years. The increase in the number of patients with associated congenital anomalies can at least partly be explained by the more sophisticated diagnostic methods we have today. The increased survival of this changed population of esophageal atresia seems to be mainly due to the improved preoperative and postoperative care and the influence of intensive care management.
The most important change in the reconstruction technique of esophageal atresia is the change from a transpleural approach to an extrapleural approach. This approach has been used since approximately 1968. Also suture materials have changed over the years from silk to polyglycolic acid sutures. In our series we have not been able to demonstrate a significant decrease in postoperative complications, ie, anastomotic leak, although the number of patients operated on before 1968 (61 patients) is probably too small to make a statistically significant difference. The number of anastomotic leaks gradually decreased over the five time periods from 11% (1947 to 1968) to 4% (1996 to 2000). Preventing anastomotic tension and pleural lesion during surgery are important to diminish the risk of postoperative complications. Our results show that anastomotic tension causes a significantly higher risk for development of anastomotic leak, and a pleural lesion causes a significantly higher risk for the development of a pneumothorax.
The patient characteristics (sex ratio, birthweight, gestational age, type of atresia, additional congenital malformations) in this series are comparable with those in other large series [5–7].
There have been multiple reports on the increased incidence of pyloric stenosis after repair of esophageal atresia [8–10]. It is not clear what the cause of this increased incidence is. Vagal nerve lesion, gastrostomy, and transpyloric feeding tubes have been mentioned as possible causes of hypertrophic pyloric stenosis after correction of esophageal atresia [8, 11]. The incidence of pyloric stenosis in the normal population is 0.1% to 1%. In our study population, a hypertrophic pyloric stenosis developed in 6% of all patients and in 13% of the patients with type A atresia.
Tracheomalacia is a common problem after correction of esophageal atresia. It can cause a typical "seal-like" cough, also referred to as TOF-cough. In more extreme cases it can be responsible for acute life threatening events. Clinically significant tracheomalacia was present in 34 of 269 patients (13%). This is consistent with the numbers (7% to 16%) that have been reported from several large series [12–14]. It is presently known that tracheomalacia can be aggravated by GER [15]. It has become our clinical experience that aggressive treatment of GER may often eradicate the symptoms and obviate the need for aortopexy.
GER is a frequently occurring problem after correction of esophageal atresia; it may cause anastomotic stricture, failure to thrive, respiratory problems, and in the long term esophagitis and intestinal metaplasia [16, 17]. Since 1986, patients have been screened routinely for GER in our center by means of pH measurements. Since then the number of patients diagnosed with GER has increased to 42%. The treatment of this GER has been more aggressive, and resulted in more gastropexies (from 15% to 32%, Table 5). As a consequence, the number of patients requiring multiple dilatations of an anastomotic stricture decreased after 1985 (10% before 1985 to 2% after 1985; Table 5; p = 0.01). The lesson learned from this experience was to interpret the occurrence of an anastomotic stricture as a symptom of GER and treatment must include antireflux therapy.
The standard antireflux procedure in our center is the Boerema anterior gastropexy. The failure rate (defined as the need for reoperation) of the Boerema anterior gastropexy in our esophageal atresia patients was 11%. This seems lower than the failure-rate of other types of fundoplication (Nissen or Thal) after esophageal atresia (15% to 30%) [18, 19].
In summary this study shows an important change in demographics in patients treated for esophageal atresia at our center: a lower birthweight, shorter gestational age, and more additional congenital anomalies, particularly cardiac anomalies. The most important cause of death has now shifted toward the mortality associated with congenital anomalies. Technical improvements of supportive care and more aggressive intensive care treatment of postoperative morbidity may have led to a further decrease of mortality after correction of esophageal atresia. It is not known what the influence of this decrease in mortality is on the (long-term) morbidity and quality of life after correction of esophageal atresia. Additional follow-up studies are in progress to learn more about the long-term outcome of the described patient population. [20]
|
References
|
|---|
-
Eerland L.D. The scalpel and the candle [Het scalpel en de kaars], 1st ed. Assen: Van Gorcum & Comp. NV, 1970.
-
Boerema I. Hiatus hernia: repair by right-sided, subhepatic, anterior gastropexy. Surgery 1969;65:884-893.[Medline]
-
Vos A., Boerema I. Surgical treatment of gastroesophageal reflux in infants and children: long-term results in 28 cases. J Pediatr Surg 1971;6:101-111.[Medline]
-
Heij H.A., Vos A. Long-term results of anterior gastropexy for gastroesophageal reflux in children. Ped Surg Int 1988;4:256-259.
-
Louhimo I., Lindahl H. Esophageal atresia: primary results of 500 consecutively treated patients. J Pediatr Surg 1983;18:217-229.[Medline]
-
Sillen U., Hagberg S., Rubenson A., Werkmaster K. Management of esophageal atresia: review of 16 years’ experience. J Pediatr Surg 1988;23:805-809.[Medline]
-
Rokitansky A.M., Kolankaya V.A., Seidl S., et al. Recent evaluation of prognostic risk factors in esophageal atresia—a multicenter review of 223 cases. Eur J Pediatr Surg 1993;3:196-201.[Medline]
-
Okada A., Usui N., Inoue M., et al. Esophageal atresia in Osaka: a review of 39 years’ experience. J Pediatr Surg 1997;32:1570-1574.[Medline]
-
Czernik J., Raine P.A. Oesophageal atresia and pyloric stenosis—an association. Z Kinderchir 1982;35:18-20.[Medline]
-
Qvist N., Rasmussen L., Hansen L.P., Pedersen S.A. Development of infantile hypertrophic pyloric stenosis in patients treated for oesophageal atresia. A case report. Acta Chir Scand 1986;152:237-238.[Medline]
-
Magilner A.D. Esophageal atresia and hypertrophic pyloric stenosis: sequential coexistence of disease. AJR Am J Roentgenol 1986;147:329-330.[Free Full Text]
-
Kilic N., Gurpinar A., Kiristioglu I., Dogruyol H. Association of oesophageal atresia and hypertrophic pyloric stenosis. Acta Paediatr 2000;89:118-119.[Medline]
-
Oldham K.T., Colombani P.M., Foglia R.P. Surgery of infants and children. New York: Lippincott-Raven, 1997.
-
Spitz L., Kiely E., Brereton R.J. Esophageal atresia: five year experience with 148 cases. J Pediatr Surg 1987;22:103-108.[Medline]
-
Tsai J.Y., Berkery L., Wesson D.E., Redo S.F., Spigland N.A. Esophageal atresia and tracheoesophageal fistula: surgical experience over two decades. Ann Thorac Surg 1997;64:778-783.[Abstract/Free Full Text]
-
Orenstein S.R., Izadnia F., Khan S. Gastroesophageal reflux disease in children. Gastroenterol Clin North Am 1999;28:947-969.[Medline]
-
Haggitt R.C. Barrett’s esophagus, dysplasia, and adenocarcinoma. Hum Pathol 1994;25:982-993.[Medline]
-
Hameeteman W., Tytgat G.N.J., Houthoff H.J., Van Den Tweel J.G. Barrett’s esophagus: development of dysplasia and adenocarcinoma. Gastroenterology 1989;96:1249-1256.[Medline]
-
Bergmeijer J.H., Tibboel D., Hazebroek F.W. Nissen fundoplication in the management of gastroesophageal reflux occuring after repair of esophageal atresia. J Pediatr Surg 2000;35:573-576.[Medline]
-
Snyder C.L., Ramachandran V., Kennedy A.P., Gittes G.K., Ashcraft K.W., Holder T.M. Efficacy of partial wrap fundoplication for gastroesophageal reflux after repair of esophageal atresia. J Pediatr Surg 1997;32:1089-1091.[Medline]
Top
Abstract
Introduction
