- Email: [email protected]
Correction of esophageal atresia with distal tracheoesophageal fistula
J THORAC
CARDIOVASC SURG
85:229-236, 1983
Correction of esophageal atresia with distal tracheoesophageal fistula Correction of the full spectrum of esophageal atresia with tracheoesophageal fistula (TEF) remains controversial. Circular myotomy and other lengthening procedures have shown promise to reduce tension when a relatively wide gap exists between esophageal segments; nevertheless a relatively high complication rate persists. We believe anastomotic tension is commonly found with repair of this anomaly. Therefore, the construction of the anastomosis will be a primary determinant of success. Twenty-four infants with TEF were admitted, 12 (50%) weighing 2.5 kg, nine (37%) 1.8 to 2.5 kg, and three (13%) 1.8 kg. All underwent gastrostomy and end-to-end single-layer anastomosis. Gaps of up to 4.5 em were encountered, and in one case a cervical incision was necessary for mobilization of the upper pouch. For eight patients (33%) the gap was at least 2.5 em and significant anastomotic tension was generated. For the series, there were no anastomotic leaks (all confirmed by barium swallow), reoperations, or surgical complications (there were two late, unrelated deaths). Prophylactic dilation was routinely performed 6 weeks and 3 months postoperatively. Subsequently, seven of the 24 (29%) required additional (one tofive) dilatations but are now asymptomatic at least 2 years later. Follow-up for the entire series is 5 months to 5 years. Three infants (13%) requiredfundoplicationfor reflux without stricture and two infants (8%) an aortopexy. For successful esophageal anastomosis we consider the following technical points important; (1) no-touch technique to minimize tissue damage, (2) generous (5 to 7 mm) full-thickness suture depth, (3) fine (6/0) monofilament suture to reduce tissue reactivity, and (4) in cases of significant tension, the sutures are preplaced and used to provide traction to eliminate tension during tying. Tension is often unavoidable in TEF, yet a carefully constructed anastomosis will withstand this stress. This approach provides results at least as satisfactory as the reported experience with a variety of techniques.
Arthur D. Santos, M.D., Theodore R. Thompson, M.D., Dana E. Johnson, M.D., Ph.D., and John E. Foker, M.D., Ph.D., Minneapolis, Minn.
T
he outlook for infants born with esophageal atresia and tracheoesophageal fistula has improved over the years, yet not all of these infants enjoy completely successful surgical correction. Even the most recent series have reported a 13% to 34% operative complication rate, with 5% to 14% having severe problems.v " The source of most of the surgical difficulties would seem to be the variable gap between the upper and lower esophageal segments. When the distance is relatively great, there will be significant anastomotic tension and, presumably, a greater chance for leak or disruption. Although this point has not been established or disproved, it is a logical assumption.
From the Departments of Surgery and Pediatrics, University of Minnesota Health Sciences Center, Minneapolis, Minn. Read at the Eighth Annual Meeting of The Samson Thoracic Surgical Society, San Diego, Calif., June 23-27, 1982. Address for reprints: John E. Foker, M.D., Box 495, University Hospitals, 420 Delaware St. S.E., Minneapolis, Minn. 55455. 0022-5223/83/020229+08$00.80/0
© 1983 The
C. V. Mosby Co.
The frequent occurrence of difficult anastomoses and the relatively constant complication rate despite the type of repair used have led several authors to conclude that the technique itself is not important. 4, 5 Anastomotic tension and, to a lesser degree, poor blood supply and a thin distal segment have been considered the major culprits. 4, 6 Consequently, the emphasis over the past few years has been on techniques designed to reduce or eliminate anastomotic tension and promote a successful repair. A variety of surgical and nonsurgical procedures have been tried and range from preliminary bougienage for pouch lengthening to circular myotomy, the technique that is enjoying considerable popularity at present.": 8 Although these techniques appear to decrease the final tension on the anastomotic line, each has been accompanied by early and late complications peculiar to the procedure. The inherent difficulties in the maneuvers designed to reduce tension have caused us to take a somewhat different course. Long gaps are often found in the spectrum of this anomaly, and our solution has been to 229
The Journal of
230 Santos et al.
Thoracic and Cardiovascular Surgery
Fig. 2. After division and closure of the fistula, stay sutures are placed to define the posterior anastomotic row. Generous tissue bites are taken in the mobilized upper pouch and the nondissected lower segment.
Fig. 1. Preoperative chest x-ray film of patient with longest esophageal gap. The catheter pushing down on the upper pouch is still just outside the thorax (white arrow). A cervical incision was required to mobilize the pouch. This infant also had a preduodenal portal vein with antral atresia, and only a stomach bubble is seen distally.
consider that some degree of tension is unavoidable. We believe that the way to produce the best overall results is to concentrate on the creation of the anastomosis rather than on techniques designed to reduce tension. Consequently, our approach for all patients with esophageal atresia and tracheoesophageal fistula has been to do a primary anastomosis without benefit of strategies to reduce tension. We acknowledged that extreme anastomotic tension will result on occasion, which violates a cardinal surgical principle. It is our feeling, however, that if certain technical details are adhered to, this violation can be overcome with satisfactory results. The purpose of this report is to examine this approach in the treatment of infants with esophageal atresia and tracheoesophageal fistula.
Patients and methods Patients. Twenty-four infants with esophageal atresia and tracheoesophageal fistula comprise the series and represent the entire operative experience of a single surgeon. The composition of the series was unremarkable. Twelve infants were male and 12 female. Twelve infants (50%) had birth weights greater than 2,500 gm, nine infants (37%) weighed between 1,800 and 2,500 gm, and three 03%) were less than 1,800 gm.
Two infants were found to have more than three components of the Vater association, which includes vertebral, radical, and renal anomalies. 9 A variety of other defects were seen, with one patient each having duodenal atresia, a preduodenal portal vein with antral atresia, ventricular septal defect, a vascular ring, malrotation, radial aplasia, and hemivertebrae. A right aortic arch was seen in two infants. Preoperative management. Diagnosis was suspected from the usual symptoms and the inability to pass a nasogastric tube. The diagnosis was confirmed either by the installation of air in the upper pouch or, in some cases, the use of nonbarium contrast medium. Fluid resuscitation and correction of electrolyte abnormalities were promptly begun. Antibiotics and vitamin K were given routinely. During the preoperative period, the infants were kept in a semiupright position and intermittent suction was applied to the proximal pouch to minimize aspiration. Eleven infants had a delayed repair, with three patients delayed beyond 20 days. Two of these infants weighed less than 1,600 gm at birth and were given parenteral nutrition until they reached 1,700 gm. Operation in the third infant was delayed until the problem of neonatal asphyxia and seizures was under control. Operative management. All infants received a Stamm gastrostomy and primary repair through a lateral thoracotomy through the fourth intercostal space on the side opposite the arch. An extrapleural approach in all cases was used. In no instance was the fistula divided and primary repair not carried out. The surgical technique was as follows. The upper pouch was located, usually with the aid of a catheter
Volume 85 Number 2
Esophageal atresia with distal TEF
23 1
February. 1983
within the lumen. A no-touch technique was used for both the upper and lower ends of the esophagus. The tissues surrounding the esophageal ends were picked up to accomplish the dissection rather than grasping the esophagus itself. Dissection was facilitated by placement of a 5-0 suture through the lowest point of the upper pouch to provide traction without the use of instruments. Because of the relatively rich submucosal blood supply, dissection of the upper pouch was generous. Particular care was taken to divide the ligamentous attachments between the upper pouch and the trachea. These were often divided between small metal clips to avoid bleeding and the use of cautery. In one case, the upper pouch was not reachable by the thoracic incision and a separate cervical incision was performed (Fig. 1). The pouch was located and dissected free so that it could be pulled down into the thorax. The lower end of the esophagus was located deep to the vagus nerve. At the point just below the inferior margin of the fistula, the esophagus was looped and the fistula dissected out. Little additional dissection of the lower end of the esophagus was carried out to preserve its blood supply. The fistula was sharply divided approximately I to 2 mm from the trachea. This allowed ample cuff for closure with interrupted 6-0 polypropylene sutures, and the repair was tested with saline. When both the upper and lower ends were dissected out an estimate of the gap could be made. For eight out of 24 (33 percent) a long (2.5 to 4.5 em) gap was present and for the remainder the gap was narrower. Direct grasping of the esophagus was avoided during the construction of the anastomosis. An incision of approximately 6 to 8 mm was made in the blind end of the upper pouch and no tissue was excised. The lower end of the esophagus was usually incised along the lateral aspect to make the lumina more nearly equal in size (Fig. 2). Care was taken to identify the mucosa within both the upper and lower segments. The most effective way we have found to visualize the mucosa without damaging the tissue is to place both legs of a fine tissue forceps within the lumen. Allowing the legs of the forceps to open will reveal the mucosal layer within the lumen. A blunt nerve hook is also helpful to elevate the tissue. Two lateral stay-stiches with very generous (5 to 7 mm) full-thickness bites of esophagus were placed (Fig. 2). The two stay-sutures were placed from inside out and served to align the ends of the two segments. The back row of interrupted 6-0 polypropylene sutures were placed from outside in so that the knots would be within the lumen. Again, very generous, full-thickness bites were taken. The entire back row was placed before the sutures were tied. If a considerable gap and tension were present, the sutures were crossed and used
Fig. 3. After placement of the posterior row of sutures, the gap is closed by applying traction to the crossed sutures (only two posterior sutures are shown for purposes of illustration). This allows individual sutures to be tied without tension.
to draw the esophageal ends together (Fig. 3). The stress was thereby distributed among the crossed sutures and allowed individual sutures to be tied without tension. The esophageal ends were brought just into apposition by tying and care was taken not to make the sutures too tight. Traction was maintained on the remaining untied sutures as the entire back row was tied one by one. The two lateral sutures were left untied so that flexibility in placing the anterior row would be present. The anterior row of sutures was placed so that the knots would be on the outside and, again, all were placed before being tied. Just before these sutures were tied, a No. 10-12 catheter was placed through the anastomosis. When all the sutures had been tied and cut, the catheter was removed. No additional reinforcing or traction sutures were used to buttress the anastomosis. Postoperative management. All infants were kept in a semiupright position with gastrostomy drainage continued. The infants were weaned as rapidly as possible from the ventilators. No deep suctioning was performed. Parenteral nutrition was maintained in all cases until an esophagogram was done 7 to 10 days later to look for both anastomotic leak and gastroesophageal reflux. At that time, the extrapleural chest tube was removed and oral feedings were begun. All infants had routine outpatient esophageal dilation with Tucker dilators to size 20 Fr. at 6 weeks and to size 24 Fr. at 3 months after repair. Additional dilations were performed as needed. The gastrostomy was removed at the time of the first clinic visit.
The Journal of
232 Santos et al.
Thoracic and Cardiovascular Surgery
Table I. Surgical therapy for 24 infants Operative results Operative related deaths Anastomotic leaks Recurrent tracheoesophageal fistula Strictures: 3-4 dilations 5-7 dilations Late problems Hospital death Unrelated death Gastroesophageal reflux which required fundoplication Tracheomalacia which required aortopexy
o
o o 7 4 3
I I 3 2
Results All infants survived the operation. One hospital death occurred in an infant who died 54 days after the operation of bronchopulmonary dysplasia, staphylococcal sepsis, and liver disease. This infant could not be weaned from the respirator. Esophagograms were done to ensure an adequate anastomosis. A second infant died after 6 months at home from the consequences of birth asphyxia and neonatal seizures. The overall survival rate was 92%. No infant had an anastomotic leak and the repair was confirmed in all cases by esophagogram (Table I). There were no recurrent tracheoesophageal fistulas. Six infants were found to have gastroesophageal reflux and three of these patients eventually required a Nissen fundoplication. The signs of mild to moderate tracheomalacia were relatively frequent and two infants underwent aortopexy to relieve tracheal compression by the innominate artery. An anastomotic stricture, as defined by the need for additional esophageal dilation, developed in seven patients (29%). Four of the seven infants received two additional dilations for a total of four and the remaining three patients had a total of five to seven dilations. All are currently asymptomatic and eating normally for age at least 1 year from the last dilation. For the entire series, all survivors are asymptomatic 4 months to 5 years after the operation except for one 2-year-old girl who occasionally continues to have the respiratory symptoms of tracheomalacia. All gastrostomy tubes have been removed and no tracts persisted and necessitated operative closure. The significant associated anomalies were also treated either prior to or at the same time as the esophageal correction. The patients with duodenal atresia and preduodenal portal vein and antral atresia had bypass procedures before the esophageal operation. The babies
with high imperforate anus and vascular ring had a colostomy performed and the ring divided at the time of esophageal repair.
Discussion The methods of treatment of esophageal atresia and tracheoesophageal fistula continue to show varied and interesting surgical approaches to the problem. Originally, several kinds of operations were tried in an attempt to achieve survival of infants with this rather complex congenital anomaly. 9 Today, with survival the norm, difficulties with the surgical repair remain and a variety of surgical solutions are used to treat these infants. Within the spectrum of tracheoesophageal fistula, it is the infants with a wide gap between the upper and lower esophageal segments who seem to pose the greatest problem. Although it may be assumed that most instances of failure occurred in patients with a long gap between segments, this has usually not been documented in clinical reports. One study, however, noted a higher stricture incidence in long-gap anastomoses." Furthermore, there is experimental evidence that the greater anastomotic tension associated with a long gap will lead to more complications.t? Because variable gaps between esophageal segments are part of the spectrum of this anomaly, in any series the repair will range from straightforward to very difficult. A variety of anastomotic techniques have been reported in the literature and include simple single-layer suturing, two-layer anastomosis, two-layer telescoping anastomosis (Haight), and the end-to-side anastomosis." Although differences for these anastomotic techniques have been claimed in terms of leak rate, stricture formation, or recurrence of tracheoesophageal fistula, in general, similar complication rates have been recorded.t- 4. 11. 12 Leaks occurred in 11% to 21% of patients, with 16.7% reported in the largest series." Strictures resulted in 25% to 45% of cases and recurrent fistulas were found in 14% of patients." A consequence of the relatively constant results has been the conclusion by some authors that the anastomotic technique itself is not important." 5 Therefore, attention has turned to ways to reduce tension prior to creation of the anastomosis. These methods, however, whether bougienage, fistulization, circular myotomy, or other techniques, all have inherent difficulties and a variety of complications have been reported with their use. Our approach has been to acknowledge that the spectrum of esophageal atresia with tracheoesophageal fistula includes the frequent occurrence of long gaps.
Volume 85 Number 2 February, 1983
Consequently, the anastomosis will also frequently be under some degree of tension. Most would agree that, to reliably succeed, an anastomosis should have the blood supply intact and be without tension. Certainly, with an impaired blood supply, an anastomosis has little chance of success. The upper esophageal pouch has a relatively rich submucosal blood supply and, therefore, can be dissected free for a considerable length. Originally, the lower limb was extensively freed up, but now the segmental and fragile nature of its blood supply is better appreciated and little dissection of the lower segment is done by most surgeons. In other anastomotic locations the importance of avoiding anastomotic tension would find general agreement; however, it is also clear that in this situation, at least, moderate degrees of tension are often encountered. Therefore, the properties of the neonatal esophagus are such that even the anastomosis under tension will frequently be successful. Our assumption has been that, if sufficient care is taken in the construction of the anastomosis, the results will be nearly uniformly good. Tension is clearly unavoidable in the repair of this anomaly and must be compensated for in the creation of the anastomosis. As outlined under the operative technique, we follow a number of details when doing the anastomosis. Although we have no evidence that anyone of them is important, there are reasons for their use. The described technique in which grasping of the esophagus is avoided should best preserve the delicate microvasculature. The suture line composed of large, full-thickness bites which include generous amounts of esophageal mucosa should be better able to withstand tension. Monofilament, fine suture of low reactivity should minimize the occurrence of microabscesses at the anastomotic line. Braided suture encourages infection and perhaps leaks in a contaminated setting." In order for the sutures to be relieved of tension during tying, traction on the remaining sutures closes the gap as described. This avoids the use of clamps and minimizes tension on the suture being tied. This maneuver and the use of very generous esophageal bites are probably the two most important details of the anastomotic techaique. We believe the size of tissue bites, at least 5 to 7 mm, deserves special emphasis, as other authors have advocated dissecting out two esophageal layers for a fine two-layer anastomosis. 14 Finally, although a single surgeon supervised the series, the repair was carried out by a wide variety of surgeons in training; therefore, the technique itself is quite practical. Several additional points about the postoperative
Esophageal atresia with distal TEF
233
treatment deserve mention. Infants in general, and babies with esophageal atresia and tracheoesophageal fistula in particular, often have significant gastroesophageal reflux. 15. 16 This tendency is combated by keeping the infants in a semiupright position and not feeding them until the contrast esophagogram is done. The gastrostomy, favored by some but not all surgeons, is performed to decompress the stomach and reduce reflux.": 9. 11 It is not used for feeding. To protect the esophageal and tracheal suture lines, deep suctioning is avoided. As with the operative technique, no proof can be offered for the value of these details, only the logic behind them given. Finally, our routine includes outpatient dilation 6 weeks and 3 months postoperatively. We have defined anastomotic stricture as the need for additional dilations, which occurred in 29% of our patients. The dilations were performed for any symptoms of swallowing difficulty and the patients were usually not investigated further with esophagograms. The strictures were very amenable to dilation and these children are now eating normally for age, at least I year after the last dilation. The three fundoplications were done for symptoms of reflux and not to reduce stricture tendency. The definition of stricture is quite variable in the literature. For those who do not routinely dilate postoperatively, any dilation requirement implies a stricture." Because we routinely dilate, we define stricture as a requirements for additional dilations. Despite the variety of definitions, the incidence is similar and ranges from 15% to 35% in most series.v v 11 Moreover, unless a major leak or anastomotic disruption occurs, or severe, persistent reflux is present, stricture is rarely a continuing problem. For the patient with a long esophageal gap, a number of solutions have been offered to facilitate the creation of the anastomosis and avoidance of tension. Bougienage has appeared in scattered reports, and most have been used for esophageal atresia without fistula.": 17. 18 Although some authors use it routinely, with a high esophageal pouch the technique seems cumbersome to us and the risk of upper pouch damage with consequent compromise of the eventual repair is present. The leak rate after its use appears to be high .19 Traction sutures either to the prevertebral fascia or the esophagus itself are probably little used today. 20 Fashioning an esophageal flap would seem to create as many problems as it solves and certainly results in a complicated suture line." Successful fistulization between pouches, either with suture alone or in combination with olives, has been reported, usually in cases of esophageal atresia
The Journal of
234 Santos et al,
without fistula. 22. 23 Although perhaps justifiable in extreme circumstances, the high risk of severe mediastinitis and the unknown long-term consequences of a relatively long stricture are not very appealing. The end-to-side anastomosis was proposed to maintain the position of the lower esophageal segment and both reduce tension and allow a larger suture line .": 24-26 Although the incidence of stricture was lessened, the leak rate was not improved and recurrent fistulization was a significant problem which in some cases necessitated reoperation. Stripping the mucosa from the fistula tract may reduce recurrence but may also be hazardous in very small infants.F The definition of long gaps in the literature is not uniform and, whereas a 1.2 em minimum may be too generous, a 2.5 em minimum certainly represents a long gap.26 A uniform and accurate measurement of gap length in the operating room is not possible and distances can represent only a best approximation. The most popular method currently used to reduce anastomotic tension is probably the circular myotomy. Livaditis" reasoned that circumferential division of the muscular layers of the upper pouch would reduce tension on the repair. The relatively rich mucosal and submucosal blood supply would preserve viability of the tissues distal to the circular myotomy. Recently, one series of five and another of nine patients with long-gap esophageal atresia and tracheoesophageal fistula have been reported. 18. 26 Although the procedure is recommended by the authors, high leak (21 %) and stricture (36%) rates were encountered. Both esophageal mucosal and tracheal injuries occurred during the procedure. Moreover, ballooning of the esophagus above the myotomy site was noted on follow-up studies. 18 The long-term significance of this remains uncertain; however, it is also clear that infants with esophageal atresia are known to have a variety of esophageal dysmotility problems.v 9, 16 Within our series, eight infants met the definition of long-gap esophageal atresia with fistula. Although it is not possible to make accurate comparisons between series, and we acknowledge that our series is small and the follow-up relatively short, it would seem that circular myotomy enjoys no obvious advantage over anastomosis without a lengthening procedure in these infants. A carefully constructed esophageal anastomosis in these infants seems able to withstand significant amounts of tension and provide intact esophageal continuity. Moreover, it should create no potential late problems and provide the greatest likelihood for longterm satisfactory function.
Thoracic and Cardiovascular Surgery
2
3
4
5
6
7
8
9
10
11
12
13
14 15
16
17
18
REFERENCES Richardson JV, Heintz SE, Rossi NP, Wright CB, Doty DB, Ehrenhaft JL: Esophageal atresia and tracheoesophageal fistula. Ann Thorac Surg 29:364-368, 1980 Hicks LM, Mansfield PB: Esophageal atresia and tracheoesophageal fistula. Review of thirteen years' experience. J THoRAc SURG 81:358-363, 1981 Holder TM, Ashcrafts KW: Development in the care of patients with esophageal atresia and tracheoesophageal fistula. Surg Clin North Am 61:1051-1061,1981 Daum R: Postoperative complications following operation for oesophageal atresia and tracheoesophageal fistula. Prog Pediatr Surg 1:209-237, 1970 Cudmore RE: Oesophageal atresia and tracheo-oesophageal fistula, Neonatal Surgery, ed 2, P Rickham, ed., London, 1978, Butterworth & Co., Ltd., pp 189-208 Schultz LR, Clatworthy HC Jr: Esophageal strictures after anastomosis in esophageal atresia. Arch Surg 87: 136140, 1963 Howard R, Myers NA: Esophageal atresia. A technique for elongating the upper pouch. Surgery 58:725-727, 1965 Livaditis A: Esophageal atresia. A method of overbridging large segmental gaps. Z Kinderchir 13:298-306, 1973 Myers NA, Aberdeen E: The esophagus, Pediatric Surgery, Vol 1, ed 3, MM Ravitch, KJ Welch, CD Benson, E Aberdeen, JG Randolph, eds., Chicago, 1979, Year Book Medical Publishers, Inc., pp 446-469 Takada Y, Kent G, Filler RM: Circular myotomy and esophageal length and safe esophageal anastomosis. An experimental study. J Pediatr Surg 16:343-348, 1981 Holder TM, Ashcraft KW: Esophageal atresia and tracheoesophageal fistula. Ann Thorac Surg 9:445-467, 1970 Pietsch JB, Stokes KB, Beardmore HE: Esophageal atresia with tracheoesophageal fistula. End-to-end versus end-to-side repair. J Pediatr Surg 13:677-681, 1978 Edlich RF, Panek PH, Rodeheaver GT, Turnbull VG, Kurtz LD, Edgerton MT: Physical and chemical configuration of sutures in the development of surgical infection. Ann Surg 177:679-688, 1973 Cloud DT: Anastomotic technique in esophageal atresia. J Pediatr Surg 3:561-564, 1968 Pieretti R, Shandling B, Stephens CA: Resistant esophageal stenosis associated with reflux after repair of esophageal atresia. A therapeutic approach. J Pediatr Surg 9:355-357, 1974 Orringer MB, Kirsh MM, Sloan H: Long-term esophageal function following repair of esophageal atresia. Ann Surg 186:436-443, 1977 Mahour GH, Woolley MM, Gwinn JL: Elongation of the upper pouch and delayed anatomic reconstruction in esophageal atresia. J Pediatr Surg 9:373-383, 1974 Ricketts RR, Luck SR, Raffensperger JG: Circular esophagomyotomy for primary repair of long-gap esophageal atresia. J Pediatr Surg 16:365-369,1981
Volume 85 Number 2
Esophageal atresia with distal TEF
February, 1983
19 Woolley MM: Esophageal atresia and tracheoesophageal fistula, 1939-1979. Am J Surg 139:771-774, 1980 20 Swenson 0: End-to-end anastomosis of the esophagus for esophageal atresia. Surgery 22:324-334, 1947 21 Kate JT: A method of suturing in operations for congenital oesophageal atresia. Arch Chir Neth 4:43-47, 1952 22 Rehbein F, Schweder N: Reconstruction of the esophagus without colon transplantation in cases of atresia. J Pediatr Surg 6:746-752, 1971 23 Shafter AD, David TE: Suture fistula as a means of connecting upper and lower segments in esophageal atresia. J Pediatr Surg 9:669-673, 1974 24 Sulamaa M, Gripenberg L, Ahvenahen EK: Prognosis and treatment of congenital atresia of the esophagus. Acta Chir Scand 102:141-157, 1951 25 Touloukian RJ, Pickett LK, Spackman T, Biancani P: Repair of esophageal atresia by end-to-side anastomosis and ligation of the tracheoesophageal fistula. A critical review of 18 cases. J Pediatr Surg 9:305-310, 1974 26 Janik JS, Filler RM, Ein SH, Simpson JS: Long-term follow-up of circular myotomy for esophageal atresia. J Pediatr Surg 15:835-841, 1980 27 Lopez-Perez GA: A modified method for the repair of esophageal atresia with tracheoesophageal fistula. Surgery 79:499-503, 1976
Discussion DR. ALAN B. GAZZANIGA Orange, Calif.
The first primary anastomosis was done by Cameron Haight in 1941. The primary repair was done through the left side of the chest with a left arch, under local anesthesia, and with extrapleural dissection. The proximal pouch was identified with a tube, much as described by the authors. Nowadays, the arch can be identified with echocardiography and the procedure done in the chest opposite the arch. In his case report, Dr. Haight gave credit to his referring physician. The patient underwent a primary repair and is alive today, at 41 or 42 years of age, and has had a normal child. Since that time, there has been a great deal of advancement in the treatment of this disorder, with excellent results. This slide shows results from Drs. Holder and Ashcraft from Kansas City. In Type A, 97% are alive from 3 to 16 years; Type B, 82%; and Type C, 50%. The authors' results 'in Type C are better than those shown here. My questions about the paper are that no time frame was given. In how long a time period were these procedures done, and how long were the follow-ups in these patients? Could the strictures that did occur have been prevented with circular myotomy? Also, using his excellent techniques in one institution, were other techniques used by other surgeons? If so, there could be comparison of the results between different surgeons,
235
DR. JOHN C. OPIE Vancouver, B.
c.. Canada
Dr. Ashmore and I do a fair amount of this work in Vancouver. My question is, what is the need for a gastrostomy? It has been mentioned rather briefly. We do not do that at all. I do not think we have done a gastrostomy for any such case unless it has been a different sort of esophageal atresia. We pass about a No.9 or No. 10 Fr. rubber catheter down to the point of the anastomosis when the anterior layer is being repaired, back it off about 1 em, and then ask the anesthetist to put a string around that at the point where it enters the nose. We send that catheter with the baby to the postoperative care unit. The nurses can then apply suction at the anastomotic suture line, to the point where the little suture is around the catheter. In that way, we perhaps reduce fistulization of the anastomosis-a small technical help. DR. LUCIUS D. HILL Seattle, Wash.
You state that in 14% of your patients a fundoplication was done, and on another slide you showed a 33% incidence of stricture. Was the stricture simply scarring at the anastomosis, or was it the result of reflux? There have been reports of patients having this kind of operation in infancy developing severe esophagitis later. Maybe you could touch on that. DR. FO KE R (Closing) I thank the discussers for their comments and I will try to take them in order. Dr. Gazzaniga, the series, we acknowledge, is a small one. The time frame is also relatively short; that is, the follow-up on these children is now between 5 months and 5 years. All of them at the present time are asymptomatic, in the sense of being able to eat whatever children of that age eat. I cannot answer whether or not a circular myotomy would have prevented stricture, and I think this overlaps the last question of Dr. Hill. Our definition of stricture was purely functional. The diagnosis was made if a child, for whatever reason, was not swallowing quite as well as he should, according to the mother. We gave that child an additional dilation, or several dilations, as an outpatient. As reported, the maximum number was five. Subsequently, all the children who had a stricture are at least 1 year after the last dilation, so they seem to do quite well with this regimen. Interestingly enough, the incidence of patients requiring additional dilations in our series, as well as the percentage requiring dilations from insitutions where prophylactic dilation is not routine, and the incidence of stricture following circular myotomy is very similar. Proponents of circular myotomy might claim that the procedure reduces the rate of stricture formation because the group in which it is would have had an even higher rate of stricturing if the lengthening procedure were not used. This has not been shown, however. The placement of the gastrostomy, we realize, is not universally agreed upon. We perform it to prevent acid reflux.
The Journal of
236 Santos et al.
The gastrostomies are left for drainage during the immediate postoperative period. We do not feed through it; we keep the children on parenteral nutrition-again, to minimize the reflux. As with many of the technical points we raised, there is just no way of knowing, without a very large comparative series from the same institution, whether these things really have any effect at all. The reason for the gastrostomy is to prevent acid reflux, which does occur, and there is a fair incidence of gastroesophageal junction incompetence in these infants. Whether it makes a difference in the late results, I cannot say. It is difficult to say to what extent continued reflux contributed to stricture formation. The Toronto group has reported that reflux can produce persistent stricturing and an antireflux procedure will reduce this tendency. Our three fundoplications were done for symptoms of regurgitation rather than for esophagitis and stricture formation. We also do not do deep suctioning. We measure the dis-
Thoracic and Cardiovascular Surgery
tance to the deep posterior pharynx and mark that on the Isolette to act as a guide for the nurses. Of course, over the series of patients additional stresses occurred. The patient with the cervical esophagus, for example, was reintubated down the esophagus twice during the course of the postoperative period. The final point is that, at our institution, this series really comprises most of the operative experience. Because there have been only a very few other patients, unfortunately we have no other comparable data. • In this report, we just wanted to raise as questions for discussion the importance of technique in creation of the anastomosis and whether or not simple anastomosis is as good as the lengthening procedures. We believe the results with simple anastomosis are as good as with those procedures. Finally, I would like to emphasize that this method is practical and easily teachable, as most of these operations were performed by residents.
CARDIOVASC SURG
85:229-236, 1983
Correction of esophageal atresia with distal tracheoesophageal fistula Correction of the full spectrum of esophageal atresia with tracheoesophageal fistula (TEF) remains controversial. Circular myotomy and other lengthening procedures have shown promise to reduce tension when a relatively wide gap exists between esophageal segments; nevertheless a relatively high complication rate persists. We believe anastomotic tension is commonly found with repair of this anomaly. Therefore, the construction of the anastomosis will be a primary determinant of success. Twenty-four infants with TEF were admitted, 12 (50%) weighing 2.5 kg, nine (37%) 1.8 to 2.5 kg, and three (13%) 1.8 kg. All underwent gastrostomy and end-to-end single-layer anastomosis. Gaps of up to 4.5 em were encountered, and in one case a cervical incision was necessary for mobilization of the upper pouch. For eight patients (33%) the gap was at least 2.5 em and significant anastomotic tension was generated. For the series, there were no anastomotic leaks (all confirmed by barium swallow), reoperations, or surgical complications (there were two late, unrelated deaths). Prophylactic dilation was routinely performed 6 weeks and 3 months postoperatively. Subsequently, seven of the 24 (29%) required additional (one tofive) dilatations but are now asymptomatic at least 2 years later. Follow-up for the entire series is 5 months to 5 years. Three infants (13%) requiredfundoplicationfor reflux without stricture and two infants (8%) an aortopexy. For successful esophageal anastomosis we consider the following technical points important; (1) no-touch technique to minimize tissue damage, (2) generous (5 to 7 mm) full-thickness suture depth, (3) fine (6/0) monofilament suture to reduce tissue reactivity, and (4) in cases of significant tension, the sutures are preplaced and used to provide traction to eliminate tension during tying. Tension is often unavoidable in TEF, yet a carefully constructed anastomosis will withstand this stress. This approach provides results at least as satisfactory as the reported experience with a variety of techniques.
Arthur D. Santos, M.D., Theodore R. Thompson, M.D., Dana E. Johnson, M.D., Ph.D., and John E. Foker, M.D., Ph.D., Minneapolis, Minn.
T
he outlook for infants born with esophageal atresia and tracheoesophageal fistula has improved over the years, yet not all of these infants enjoy completely successful surgical correction. Even the most recent series have reported a 13% to 34% operative complication rate, with 5% to 14% having severe problems.v " The source of most of the surgical difficulties would seem to be the variable gap between the upper and lower esophageal segments. When the distance is relatively great, there will be significant anastomotic tension and, presumably, a greater chance for leak or disruption. Although this point has not been established or disproved, it is a logical assumption.
From the Departments of Surgery and Pediatrics, University of Minnesota Health Sciences Center, Minneapolis, Minn. Read at the Eighth Annual Meeting of The Samson Thoracic Surgical Society, San Diego, Calif., June 23-27, 1982. Address for reprints: John E. Foker, M.D., Box 495, University Hospitals, 420 Delaware St. S.E., Minneapolis, Minn. 55455. 0022-5223/83/020229+08$00.80/0
© 1983 The
C. V. Mosby Co.
The frequent occurrence of difficult anastomoses and the relatively constant complication rate despite the type of repair used have led several authors to conclude that the technique itself is not important. 4, 5 Anastomotic tension and, to a lesser degree, poor blood supply and a thin distal segment have been considered the major culprits. 4, 6 Consequently, the emphasis over the past few years has been on techniques designed to reduce or eliminate anastomotic tension and promote a successful repair. A variety of surgical and nonsurgical procedures have been tried and range from preliminary bougienage for pouch lengthening to circular myotomy, the technique that is enjoying considerable popularity at present.": 8 Although these techniques appear to decrease the final tension on the anastomotic line, each has been accompanied by early and late complications peculiar to the procedure. The inherent difficulties in the maneuvers designed to reduce tension have caused us to take a somewhat different course. Long gaps are often found in the spectrum of this anomaly, and our solution has been to 229
The Journal of
230 Santos et al.
Thoracic and Cardiovascular Surgery
Fig. 2. After division and closure of the fistula, stay sutures are placed to define the posterior anastomotic row. Generous tissue bites are taken in the mobilized upper pouch and the nondissected lower segment.
Fig. 1. Preoperative chest x-ray film of patient with longest esophageal gap. The catheter pushing down on the upper pouch is still just outside the thorax (white arrow). A cervical incision was required to mobilize the pouch. This infant also had a preduodenal portal vein with antral atresia, and only a stomach bubble is seen distally.
consider that some degree of tension is unavoidable. We believe that the way to produce the best overall results is to concentrate on the creation of the anastomosis rather than on techniques designed to reduce tension. Consequently, our approach for all patients with esophageal atresia and tracheoesophageal fistula has been to do a primary anastomosis without benefit of strategies to reduce tension. We acknowledged that extreme anastomotic tension will result on occasion, which violates a cardinal surgical principle. It is our feeling, however, that if certain technical details are adhered to, this violation can be overcome with satisfactory results. The purpose of this report is to examine this approach in the treatment of infants with esophageal atresia and tracheoesophageal fistula.
Patients and methods Patients. Twenty-four infants with esophageal atresia and tracheoesophageal fistula comprise the series and represent the entire operative experience of a single surgeon. The composition of the series was unremarkable. Twelve infants were male and 12 female. Twelve infants (50%) had birth weights greater than 2,500 gm, nine infants (37%) weighed between 1,800 and 2,500 gm, and three 03%) were less than 1,800 gm.
Two infants were found to have more than three components of the Vater association, which includes vertebral, radical, and renal anomalies. 9 A variety of other defects were seen, with one patient each having duodenal atresia, a preduodenal portal vein with antral atresia, ventricular septal defect, a vascular ring, malrotation, radial aplasia, and hemivertebrae. A right aortic arch was seen in two infants. Preoperative management. Diagnosis was suspected from the usual symptoms and the inability to pass a nasogastric tube. The diagnosis was confirmed either by the installation of air in the upper pouch or, in some cases, the use of nonbarium contrast medium. Fluid resuscitation and correction of electrolyte abnormalities were promptly begun. Antibiotics and vitamin K were given routinely. During the preoperative period, the infants were kept in a semiupright position and intermittent suction was applied to the proximal pouch to minimize aspiration. Eleven infants had a delayed repair, with three patients delayed beyond 20 days. Two of these infants weighed less than 1,600 gm at birth and were given parenteral nutrition until they reached 1,700 gm. Operation in the third infant was delayed until the problem of neonatal asphyxia and seizures was under control. Operative management. All infants received a Stamm gastrostomy and primary repair through a lateral thoracotomy through the fourth intercostal space on the side opposite the arch. An extrapleural approach in all cases was used. In no instance was the fistula divided and primary repair not carried out. The surgical technique was as follows. The upper pouch was located, usually with the aid of a catheter
Volume 85 Number 2
Esophageal atresia with distal TEF
23 1
February. 1983
within the lumen. A no-touch technique was used for both the upper and lower ends of the esophagus. The tissues surrounding the esophageal ends were picked up to accomplish the dissection rather than grasping the esophagus itself. Dissection was facilitated by placement of a 5-0 suture through the lowest point of the upper pouch to provide traction without the use of instruments. Because of the relatively rich submucosal blood supply, dissection of the upper pouch was generous. Particular care was taken to divide the ligamentous attachments between the upper pouch and the trachea. These were often divided between small metal clips to avoid bleeding and the use of cautery. In one case, the upper pouch was not reachable by the thoracic incision and a separate cervical incision was performed (Fig. 1). The pouch was located and dissected free so that it could be pulled down into the thorax. The lower end of the esophagus was located deep to the vagus nerve. At the point just below the inferior margin of the fistula, the esophagus was looped and the fistula dissected out. Little additional dissection of the lower end of the esophagus was carried out to preserve its blood supply. The fistula was sharply divided approximately I to 2 mm from the trachea. This allowed ample cuff for closure with interrupted 6-0 polypropylene sutures, and the repair was tested with saline. When both the upper and lower ends were dissected out an estimate of the gap could be made. For eight out of 24 (33 percent) a long (2.5 to 4.5 em) gap was present and for the remainder the gap was narrower. Direct grasping of the esophagus was avoided during the construction of the anastomosis. An incision of approximately 6 to 8 mm was made in the blind end of the upper pouch and no tissue was excised. The lower end of the esophagus was usually incised along the lateral aspect to make the lumina more nearly equal in size (Fig. 2). Care was taken to identify the mucosa within both the upper and lower segments. The most effective way we have found to visualize the mucosa without damaging the tissue is to place both legs of a fine tissue forceps within the lumen. Allowing the legs of the forceps to open will reveal the mucosal layer within the lumen. A blunt nerve hook is also helpful to elevate the tissue. Two lateral stay-stiches with very generous (5 to 7 mm) full-thickness bites of esophagus were placed (Fig. 2). The two stay-sutures were placed from inside out and served to align the ends of the two segments. The back row of interrupted 6-0 polypropylene sutures were placed from outside in so that the knots would be within the lumen. Again, very generous, full-thickness bites were taken. The entire back row was placed before the sutures were tied. If a considerable gap and tension were present, the sutures were crossed and used
Fig. 3. After placement of the posterior row of sutures, the gap is closed by applying traction to the crossed sutures (only two posterior sutures are shown for purposes of illustration). This allows individual sutures to be tied without tension.
to draw the esophageal ends together (Fig. 3). The stress was thereby distributed among the crossed sutures and allowed individual sutures to be tied without tension. The esophageal ends were brought just into apposition by tying and care was taken not to make the sutures too tight. Traction was maintained on the remaining untied sutures as the entire back row was tied one by one. The two lateral sutures were left untied so that flexibility in placing the anterior row would be present. The anterior row of sutures was placed so that the knots would be on the outside and, again, all were placed before being tied. Just before these sutures were tied, a No. 10-12 catheter was placed through the anastomosis. When all the sutures had been tied and cut, the catheter was removed. No additional reinforcing or traction sutures were used to buttress the anastomosis. Postoperative management. All infants were kept in a semiupright position with gastrostomy drainage continued. The infants were weaned as rapidly as possible from the ventilators. No deep suctioning was performed. Parenteral nutrition was maintained in all cases until an esophagogram was done 7 to 10 days later to look for both anastomotic leak and gastroesophageal reflux. At that time, the extrapleural chest tube was removed and oral feedings were begun. All infants had routine outpatient esophageal dilation with Tucker dilators to size 20 Fr. at 6 weeks and to size 24 Fr. at 3 months after repair. Additional dilations were performed as needed. The gastrostomy was removed at the time of the first clinic visit.
The Journal of
232 Santos et al.
Thoracic and Cardiovascular Surgery
Table I. Surgical therapy for 24 infants Operative results Operative related deaths Anastomotic leaks Recurrent tracheoesophageal fistula Strictures: 3-4 dilations 5-7 dilations Late problems Hospital death Unrelated death Gastroesophageal reflux which required fundoplication Tracheomalacia which required aortopexy
o
o o 7 4 3
I I 3 2
Results All infants survived the operation. One hospital death occurred in an infant who died 54 days after the operation of bronchopulmonary dysplasia, staphylococcal sepsis, and liver disease. This infant could not be weaned from the respirator. Esophagograms were done to ensure an adequate anastomosis. A second infant died after 6 months at home from the consequences of birth asphyxia and neonatal seizures. The overall survival rate was 92%. No infant had an anastomotic leak and the repair was confirmed in all cases by esophagogram (Table I). There were no recurrent tracheoesophageal fistulas. Six infants were found to have gastroesophageal reflux and three of these patients eventually required a Nissen fundoplication. The signs of mild to moderate tracheomalacia were relatively frequent and two infants underwent aortopexy to relieve tracheal compression by the innominate artery. An anastomotic stricture, as defined by the need for additional esophageal dilation, developed in seven patients (29%). Four of the seven infants received two additional dilations for a total of four and the remaining three patients had a total of five to seven dilations. All are currently asymptomatic and eating normally for age at least 1 year from the last dilation. For the entire series, all survivors are asymptomatic 4 months to 5 years after the operation except for one 2-year-old girl who occasionally continues to have the respiratory symptoms of tracheomalacia. All gastrostomy tubes have been removed and no tracts persisted and necessitated operative closure. The significant associated anomalies were also treated either prior to or at the same time as the esophageal correction. The patients with duodenal atresia and preduodenal portal vein and antral atresia had bypass procedures before the esophageal operation. The babies
with high imperforate anus and vascular ring had a colostomy performed and the ring divided at the time of esophageal repair.
Discussion The methods of treatment of esophageal atresia and tracheoesophageal fistula continue to show varied and interesting surgical approaches to the problem. Originally, several kinds of operations were tried in an attempt to achieve survival of infants with this rather complex congenital anomaly. 9 Today, with survival the norm, difficulties with the surgical repair remain and a variety of surgical solutions are used to treat these infants. Within the spectrum of tracheoesophageal fistula, it is the infants with a wide gap between the upper and lower esophageal segments who seem to pose the greatest problem. Although it may be assumed that most instances of failure occurred in patients with a long gap between segments, this has usually not been documented in clinical reports. One study, however, noted a higher stricture incidence in long-gap anastomoses." Furthermore, there is experimental evidence that the greater anastomotic tension associated with a long gap will lead to more complications.t? Because variable gaps between esophageal segments are part of the spectrum of this anomaly, in any series the repair will range from straightforward to very difficult. A variety of anastomotic techniques have been reported in the literature and include simple single-layer suturing, two-layer anastomosis, two-layer telescoping anastomosis (Haight), and the end-to-side anastomosis." Although differences for these anastomotic techniques have been claimed in terms of leak rate, stricture formation, or recurrence of tracheoesophageal fistula, in general, similar complication rates have been recorded.t- 4. 11. 12 Leaks occurred in 11% to 21% of patients, with 16.7% reported in the largest series." Strictures resulted in 25% to 45% of cases and recurrent fistulas were found in 14% of patients." A consequence of the relatively constant results has been the conclusion by some authors that the anastomotic technique itself is not important." 5 Therefore, attention has turned to ways to reduce tension prior to creation of the anastomosis. These methods, however, whether bougienage, fistulization, circular myotomy, or other techniques, all have inherent difficulties and a variety of complications have been reported with their use. Our approach has been to acknowledge that the spectrum of esophageal atresia with tracheoesophageal fistula includes the frequent occurrence of long gaps.
Volume 85 Number 2 February, 1983
Consequently, the anastomosis will also frequently be under some degree of tension. Most would agree that, to reliably succeed, an anastomosis should have the blood supply intact and be without tension. Certainly, with an impaired blood supply, an anastomosis has little chance of success. The upper esophageal pouch has a relatively rich submucosal blood supply and, therefore, can be dissected free for a considerable length. Originally, the lower limb was extensively freed up, but now the segmental and fragile nature of its blood supply is better appreciated and little dissection of the lower segment is done by most surgeons. In other anastomotic locations the importance of avoiding anastomotic tension would find general agreement; however, it is also clear that in this situation, at least, moderate degrees of tension are often encountered. Therefore, the properties of the neonatal esophagus are such that even the anastomosis under tension will frequently be successful. Our assumption has been that, if sufficient care is taken in the construction of the anastomosis, the results will be nearly uniformly good. Tension is clearly unavoidable in the repair of this anomaly and must be compensated for in the creation of the anastomosis. As outlined under the operative technique, we follow a number of details when doing the anastomosis. Although we have no evidence that anyone of them is important, there are reasons for their use. The described technique in which grasping of the esophagus is avoided should best preserve the delicate microvasculature. The suture line composed of large, full-thickness bites which include generous amounts of esophageal mucosa should be better able to withstand tension. Monofilament, fine suture of low reactivity should minimize the occurrence of microabscesses at the anastomotic line. Braided suture encourages infection and perhaps leaks in a contaminated setting." In order for the sutures to be relieved of tension during tying, traction on the remaining sutures closes the gap as described. This avoids the use of clamps and minimizes tension on the suture being tied. This maneuver and the use of very generous esophageal bites are probably the two most important details of the anastomotic techaique. We believe the size of tissue bites, at least 5 to 7 mm, deserves special emphasis, as other authors have advocated dissecting out two esophageal layers for a fine two-layer anastomosis. 14 Finally, although a single surgeon supervised the series, the repair was carried out by a wide variety of surgeons in training; therefore, the technique itself is quite practical. Several additional points about the postoperative
Esophageal atresia with distal TEF
233
treatment deserve mention. Infants in general, and babies with esophageal atresia and tracheoesophageal fistula in particular, often have significant gastroesophageal reflux. 15. 16 This tendency is combated by keeping the infants in a semiupright position and not feeding them until the contrast esophagogram is done. The gastrostomy, favored by some but not all surgeons, is performed to decompress the stomach and reduce reflux.": 9. 11 It is not used for feeding. To protect the esophageal and tracheal suture lines, deep suctioning is avoided. As with the operative technique, no proof can be offered for the value of these details, only the logic behind them given. Finally, our routine includes outpatient dilation 6 weeks and 3 months postoperatively. We have defined anastomotic stricture as the need for additional dilations, which occurred in 29% of our patients. The dilations were performed for any symptoms of swallowing difficulty and the patients were usually not investigated further with esophagograms. The strictures were very amenable to dilation and these children are now eating normally for age, at least I year after the last dilation. The three fundoplications were done for symptoms of reflux and not to reduce stricture tendency. The definition of stricture is quite variable in the literature. For those who do not routinely dilate postoperatively, any dilation requirement implies a stricture." Because we routinely dilate, we define stricture as a requirements for additional dilations. Despite the variety of definitions, the incidence is similar and ranges from 15% to 35% in most series.v v 11 Moreover, unless a major leak or anastomotic disruption occurs, or severe, persistent reflux is present, stricture is rarely a continuing problem. For the patient with a long esophageal gap, a number of solutions have been offered to facilitate the creation of the anastomosis and avoidance of tension. Bougienage has appeared in scattered reports, and most have been used for esophageal atresia without fistula.": 17. 18 Although some authors use it routinely, with a high esophageal pouch the technique seems cumbersome to us and the risk of upper pouch damage with consequent compromise of the eventual repair is present. The leak rate after its use appears to be high .19 Traction sutures either to the prevertebral fascia or the esophagus itself are probably little used today. 20 Fashioning an esophageal flap would seem to create as many problems as it solves and certainly results in a complicated suture line." Successful fistulization between pouches, either with suture alone or in combination with olives, has been reported, usually in cases of esophageal atresia
The Journal of
234 Santos et al,
without fistula. 22. 23 Although perhaps justifiable in extreme circumstances, the high risk of severe mediastinitis and the unknown long-term consequences of a relatively long stricture are not very appealing. The end-to-side anastomosis was proposed to maintain the position of the lower esophageal segment and both reduce tension and allow a larger suture line .": 24-26 Although the incidence of stricture was lessened, the leak rate was not improved and recurrent fistulization was a significant problem which in some cases necessitated reoperation. Stripping the mucosa from the fistula tract may reduce recurrence but may also be hazardous in very small infants.F The definition of long gaps in the literature is not uniform and, whereas a 1.2 em minimum may be too generous, a 2.5 em minimum certainly represents a long gap.26 A uniform and accurate measurement of gap length in the operating room is not possible and distances can represent only a best approximation. The most popular method currently used to reduce anastomotic tension is probably the circular myotomy. Livaditis" reasoned that circumferential division of the muscular layers of the upper pouch would reduce tension on the repair. The relatively rich mucosal and submucosal blood supply would preserve viability of the tissues distal to the circular myotomy. Recently, one series of five and another of nine patients with long-gap esophageal atresia and tracheoesophageal fistula have been reported. 18. 26 Although the procedure is recommended by the authors, high leak (21 %) and stricture (36%) rates were encountered. Both esophageal mucosal and tracheal injuries occurred during the procedure. Moreover, ballooning of the esophagus above the myotomy site was noted on follow-up studies. 18 The long-term significance of this remains uncertain; however, it is also clear that infants with esophageal atresia are known to have a variety of esophageal dysmotility problems.v 9, 16 Within our series, eight infants met the definition of long-gap esophageal atresia with fistula. Although it is not possible to make accurate comparisons between series, and we acknowledge that our series is small and the follow-up relatively short, it would seem that circular myotomy enjoys no obvious advantage over anastomosis without a lengthening procedure in these infants. A carefully constructed esophageal anastomosis in these infants seems able to withstand significant amounts of tension and provide intact esophageal continuity. Moreover, it should create no potential late problems and provide the greatest likelihood for longterm satisfactory function.
Thoracic and Cardiovascular Surgery
2
3
4
5
6
7
8
9
10
11
12
13
14 15
16
17
18
REFERENCES Richardson JV, Heintz SE, Rossi NP, Wright CB, Doty DB, Ehrenhaft JL: Esophageal atresia and tracheoesophageal fistula. Ann Thorac Surg 29:364-368, 1980 Hicks LM, Mansfield PB: Esophageal atresia and tracheoesophageal fistula. Review of thirteen years' experience. J THoRAc SURG 81:358-363, 1981 Holder TM, Ashcrafts KW: Development in the care of patients with esophageal atresia and tracheoesophageal fistula. Surg Clin North Am 61:1051-1061,1981 Daum R: Postoperative complications following operation for oesophageal atresia and tracheoesophageal fistula. Prog Pediatr Surg 1:209-237, 1970 Cudmore RE: Oesophageal atresia and tracheo-oesophageal fistula, Neonatal Surgery, ed 2, P Rickham, ed., London, 1978, Butterworth & Co., Ltd., pp 189-208 Schultz LR, Clatworthy HC Jr: Esophageal strictures after anastomosis in esophageal atresia. Arch Surg 87: 136140, 1963 Howard R, Myers NA: Esophageal atresia. A technique for elongating the upper pouch. Surgery 58:725-727, 1965 Livaditis A: Esophageal atresia. A method of overbridging large segmental gaps. Z Kinderchir 13:298-306, 1973 Myers NA, Aberdeen E: The esophagus, Pediatric Surgery, Vol 1, ed 3, MM Ravitch, KJ Welch, CD Benson, E Aberdeen, JG Randolph, eds., Chicago, 1979, Year Book Medical Publishers, Inc., pp 446-469 Takada Y, Kent G, Filler RM: Circular myotomy and esophageal length and safe esophageal anastomosis. An experimental study. J Pediatr Surg 16:343-348, 1981 Holder TM, Ashcraft KW: Esophageal atresia and tracheoesophageal fistula. Ann Thorac Surg 9:445-467, 1970 Pietsch JB, Stokes KB, Beardmore HE: Esophageal atresia with tracheoesophageal fistula. End-to-end versus end-to-side repair. J Pediatr Surg 13:677-681, 1978 Edlich RF, Panek PH, Rodeheaver GT, Turnbull VG, Kurtz LD, Edgerton MT: Physical and chemical configuration of sutures in the development of surgical infection. Ann Surg 177:679-688, 1973 Cloud DT: Anastomotic technique in esophageal atresia. J Pediatr Surg 3:561-564, 1968 Pieretti R, Shandling B, Stephens CA: Resistant esophageal stenosis associated with reflux after repair of esophageal atresia. A therapeutic approach. J Pediatr Surg 9:355-357, 1974 Orringer MB, Kirsh MM, Sloan H: Long-term esophageal function following repair of esophageal atresia. Ann Surg 186:436-443, 1977 Mahour GH, Woolley MM, Gwinn JL: Elongation of the upper pouch and delayed anatomic reconstruction in esophageal atresia. J Pediatr Surg 9:373-383, 1974 Ricketts RR, Luck SR, Raffensperger JG: Circular esophagomyotomy for primary repair of long-gap esophageal atresia. J Pediatr Surg 16:365-369,1981
Volume 85 Number 2
Esophageal atresia with distal TEF
February, 1983
19 Woolley MM: Esophageal atresia and tracheoesophageal fistula, 1939-1979. Am J Surg 139:771-774, 1980 20 Swenson 0: End-to-end anastomosis of the esophagus for esophageal atresia. Surgery 22:324-334, 1947 21 Kate JT: A method of suturing in operations for congenital oesophageal atresia. Arch Chir Neth 4:43-47, 1952 22 Rehbein F, Schweder N: Reconstruction of the esophagus without colon transplantation in cases of atresia. J Pediatr Surg 6:746-752, 1971 23 Shafter AD, David TE: Suture fistula as a means of connecting upper and lower segments in esophageal atresia. J Pediatr Surg 9:669-673, 1974 24 Sulamaa M, Gripenberg L, Ahvenahen EK: Prognosis and treatment of congenital atresia of the esophagus. Acta Chir Scand 102:141-157, 1951 25 Touloukian RJ, Pickett LK, Spackman T, Biancani P: Repair of esophageal atresia by end-to-side anastomosis and ligation of the tracheoesophageal fistula. A critical review of 18 cases. J Pediatr Surg 9:305-310, 1974 26 Janik JS, Filler RM, Ein SH, Simpson JS: Long-term follow-up of circular myotomy for esophageal atresia. J Pediatr Surg 15:835-841, 1980 27 Lopez-Perez GA: A modified method for the repair of esophageal atresia with tracheoesophageal fistula. Surgery 79:499-503, 1976
Discussion DR. ALAN B. GAZZANIGA Orange, Calif.
The first primary anastomosis was done by Cameron Haight in 1941. The primary repair was done through the left side of the chest with a left arch, under local anesthesia, and with extrapleural dissection. The proximal pouch was identified with a tube, much as described by the authors. Nowadays, the arch can be identified with echocardiography and the procedure done in the chest opposite the arch. In his case report, Dr. Haight gave credit to his referring physician. The patient underwent a primary repair and is alive today, at 41 or 42 years of age, and has had a normal child. Since that time, there has been a great deal of advancement in the treatment of this disorder, with excellent results. This slide shows results from Drs. Holder and Ashcraft from Kansas City. In Type A, 97% are alive from 3 to 16 years; Type B, 82%; and Type C, 50%. The authors' results 'in Type C are better than those shown here. My questions about the paper are that no time frame was given. In how long a time period were these procedures done, and how long were the follow-ups in these patients? Could the strictures that did occur have been prevented with circular myotomy? Also, using his excellent techniques in one institution, were other techniques used by other surgeons? If so, there could be comparison of the results between different surgeons,
235
DR. JOHN C. OPIE Vancouver, B.
c.. Canada
Dr. Ashmore and I do a fair amount of this work in Vancouver. My question is, what is the need for a gastrostomy? It has been mentioned rather briefly. We do not do that at all. I do not think we have done a gastrostomy for any such case unless it has been a different sort of esophageal atresia. We pass about a No.9 or No. 10 Fr. rubber catheter down to the point of the anastomosis when the anterior layer is being repaired, back it off about 1 em, and then ask the anesthetist to put a string around that at the point where it enters the nose. We send that catheter with the baby to the postoperative care unit. The nurses can then apply suction at the anastomotic suture line, to the point where the little suture is around the catheter. In that way, we perhaps reduce fistulization of the anastomosis-a small technical help. DR. LUCIUS D. HILL Seattle, Wash.
You state that in 14% of your patients a fundoplication was done, and on another slide you showed a 33% incidence of stricture. Was the stricture simply scarring at the anastomosis, or was it the result of reflux? There have been reports of patients having this kind of operation in infancy developing severe esophagitis later. Maybe you could touch on that. DR. FO KE R (Closing) I thank the discussers for their comments and I will try to take them in order. Dr. Gazzaniga, the series, we acknowledge, is a small one. The time frame is also relatively short; that is, the follow-up on these children is now between 5 months and 5 years. All of them at the present time are asymptomatic, in the sense of being able to eat whatever children of that age eat. I cannot answer whether or not a circular myotomy would have prevented stricture, and I think this overlaps the last question of Dr. Hill. Our definition of stricture was purely functional. The diagnosis was made if a child, for whatever reason, was not swallowing quite as well as he should, according to the mother. We gave that child an additional dilation, or several dilations, as an outpatient. As reported, the maximum number was five. Subsequently, all the children who had a stricture are at least 1 year after the last dilation, so they seem to do quite well with this regimen. Interestingly enough, the incidence of patients requiring additional dilations in our series, as well as the percentage requiring dilations from insitutions where prophylactic dilation is not routine, and the incidence of stricture following circular myotomy is very similar. Proponents of circular myotomy might claim that the procedure reduces the rate of stricture formation because the group in which it is would have had an even higher rate of stricturing if the lengthening procedure were not used. This has not been shown, however. The placement of the gastrostomy, we realize, is not universally agreed upon. We perform it to prevent acid reflux.
The Journal of
236 Santos et al.
The gastrostomies are left for drainage during the immediate postoperative period. We do not feed through it; we keep the children on parenteral nutrition-again, to minimize the reflux. As with many of the technical points we raised, there is just no way of knowing, without a very large comparative series from the same institution, whether these things really have any effect at all. The reason for the gastrostomy is to prevent acid reflux, which does occur, and there is a fair incidence of gastroesophageal junction incompetence in these infants. Whether it makes a difference in the late results, I cannot say. It is difficult to say to what extent continued reflux contributed to stricture formation. The Toronto group has reported that reflux can produce persistent stricturing and an antireflux procedure will reduce this tendency. Our three fundoplications were done for symptoms of regurgitation rather than for esophagitis and stricture formation. We also do not do deep suctioning. We measure the dis-
Thoracic and Cardiovascular Surgery
tance to the deep posterior pharynx and mark that on the Isolette to act as a guide for the nurses. Of course, over the series of patients additional stresses occurred. The patient with the cervical esophagus, for example, was reintubated down the esophagus twice during the course of the postoperative period. The final point is that, at our institution, this series really comprises most of the operative experience. Because there have been only a very few other patients, unfortunately we have no other comparable data. • In this report, we just wanted to raise as questions for discussion the importance of technique in creation of the anastomosis and whether or not simple anastomosis is as good as the lengthening procedures. We believe the results with simple anastomosis are as good as with those procedures. Finally, I would like to emphasize that this method is practical and easily teachable, as most of these operations were performed by residents.