Management of esophageal atresia and tracheoesophageal fistula in the neonate with severe respiratory distress syndrome

Management of Esophageal Atresia and Tracheoesophageal Fistula in the Neonate With Severe Respiratory Distress Syndrome By John M. Templeton, Jr, Josephine J. Templeton, Louise Schnaufer, Harry C. Bishop, Moritz M. Ziegler, and James A. O'Neill, Jr Philadelphia 9 In a lO-year period, 22 neonates with esophageal atresia (EA) and tracheoesophageal fistula (TEF) required high pressure ventilatory support soon after birth because of respiratory distress syndrome (RDS). Eleven of the 22 or 5 0 % survived overall, but if the 5 patients w h o died before definitive surgical repair could be attempted are excluded, 11 of 17 or 6 5 % survived. M o r e importantly, 4 of 7 (57%) patients w h o had gastrostomy performed first survived while 7 of 10 (70%) w h o had fistula ligation performed first survived. The difficulties with intraoperative management of those w h o had gastrostomy performed first w e r e even more impressive. Our experience leads us to conclude that patients with EA and TEF with severe RDS w h o require high pressure ventilation preoperatively represent a group of patients w h o require special consideration. The danger to such patients with increased pulmonary resistance is not gastric distention but sudden loss of intragastric pressure. In the presence of poor lung compliance, the upper gastrointestinal tract functions in continuity with the tracheobronchial tree. A sudden loss of intragastric pressure, as with placement of a gastrostomy tube, results in an acute loss of effective ventilating pressure. Resuscitation of such a patient is not possible until leakage from the esophagus is controlled by ligation of the fistula or transabdominal occlusion of the distal esophagus. Placement of a Fogarty catheter into the fistula via a bronchoscope is effective but may not be feasible in every case. Early thoracotomy and ligation of the fistula in patients with progressive RDS provides immediate improvement in ventilatory efficiency and relief of gastric distention. 9 1 9 8 5 by Grune a n d S t r a t t o n . Inc,

gastrostomy followed by a thoracotomy at a later time when the patient's condition improved. 2-5 We have found that this approach has often resulted in acute and catastrophic cardiorespiratory deterioration in the child with RDS who requires high pressure ventilatory support. This report reviews our experience with a group of 22 patients with EA and T E F seen over the last 10 years who required ventilatory support soon after birth because of respiratory failure.

INDEX W O R D S : Esophageal atresia; tracheoesophageal fistula; respiratory distress syndrome.

A 1.36 kg female infant with EA and TEF was intubated and ventilated because of increasing respiratory distress due to RDS. At that time, PaO2 was 82 torr, PCO2 46, and pH 7.34 on a FIO2 of 80%. Operation was delayed to allow the respiratory distress to improve, but progressive abdominal distention and further clinical deterioration occurred. Emergency gastrostomy was undertaken. Prior to operation, PaO2 was 58, PCO2 59, and pH 7.26 on FIO2 of 100%. As the stomach was vented, the anesthesiologist noted sudden inability to ventilate adequately. The patient became progressively bradycardic necessitating external cardiac massage. An arterial blood gas on a FIO2 of 100% now showed a PaO: of 31 torr. Ventilation was still inadequate until digital pressure was applied to the gastroesophageal junction from within the abdomen. Ventilation them improved and the patient's pulse returned to normal. Finger compression of the lower esophagus was maintained while the patient was positioned laterally and the TEF ligated via a right thoracotomy. Followed ligation of the fistula, the P~O2 increased to 96. By 12 hours after surgery, she showed further improvement with a P~O2 of 80 on an FIO: of 58%.

T U R I T Y , serious associated anomalies, p RandE M Arespiratory distress syndrome (RDS) are factors that adversely affect survival in infants with esophageal atresia (EA) and tracheoesophageal fistula (TEF). Attempts at primary repair in such patients has resulted in mortality rates as high as 64%. 1 Earlier reported experience indicated that patients with one or more such conditions should be staged by means of a From the Department of Pediatric Surgery, University of Pennsylvania School of Medicine and The Children's Hospital of Philadelphia, Presented at the 33rd Annual Meeting of the Surgical Section of the American Academy of Pediatrics, Chicago, September 15-16, 1984. Address reprint requests to John M. Templeton, Jr, MD, Division of Pediatric Surgery, The Children's Hospital of Philadelphia, 34th St and Civic Center Blvd, Philadelphia, PA 19104. 9 1985 by Grune & Stratton, Inc. 0022-3468/85/2004-0020503.00/0 394

CLINICAL MATERIAL In a 10-year period from 1974 through 1983, 82 neonates were admitted to the Children's Hospital of Philadelphia for definitive treatment of esophageal atresia. Seventy-seven patients had EA with TEF. Of the 77 patients, 22 patients developed progressive respiratory distress necessitating preoperative ventilatory support (Table 1). One patient was managed preoperatively by continuous nasal positive airway pressure (CPAP). The other 21 patients all required endotracheal intubation and positive pressure ventilation soon after birth. Nine of the 22 patients were seen during the period 1974 to 1979. Two died before operation could be undertaken. Two of the remaining seven patients had division of the TEF with esophageal anastomosis before undergoing a gastrostomy without any intraoperative difficulties. The other five patients had a gastrostomy performed first, and four of them developed a persistent loss of air via the gastrostomy tube followed by an immediate, marked deterioration of their ventilatory status. The fifth patient did not have this problem.

Illustrative Case Report

COMMENTS B e c a u s e o f t h i s e x p e r i e n c e in e a r l y 1979, o u r a p p r o a c h g e n e r a l l y c h a n g e d in t h e m a n a g e m e n t o f those newborns with EA and TEF who required high Journal of Pediatric Surgery, Vol 20, No 4 (August), 1985: pp 394-397

MANAGEMENT OF EA AND TEF

395

Table 1. Respiratory Distress Syndrome and Distal

Name Patients

Initial ' Surgical Approach (Gast. or TEF)

Birth Weight (kg)

Dying

Tracheoesophageal Fistula*

Deteriorated Under Anesthesia

Primary Repair During 1st Op.

Survived

Associated Disorders

Before Repair

BGY JZ TS BBS BGC Patients Who

1,01 Preop death 1.00 Preop death 1.12 Preop death t 2,44 Preop death 1.08 Preop death Underwent a Gastrostomy First

------

NA NA No NA NA

No No No No No

Hypoplasia of lung, heart, & bowel

BW BBC

1,43 3.07

Gast. Gast.

Yes Yes

No Yes

Yes No

Hydrocephalus and PDA TGA, VSD

RA RS KB KS

3.01 1.86 1,36 3.06

Gast. Gast. Gast. Gast.

Yes No Yes Yes

No Yes No Yes

No Yes No Yes

Trisomy 18 Tracheomalacia IVH None

JG 1.48 Gast. Patients Who Had Ligation of TEF First JT 1.10 TEF

Yes

No

Yes

None

No

No

No

NEC

SG GT WB BGC

3.10 1.35 2.38 2.00

TEF TEF TEF TEF

No No No No

Yes No Yes No

Yes No Yes No

None Tetralogy of Fallot Recurrent TEF PFC

RH MH BH

1.83 3.80 2.70

TEF TEF TEF

No No No

No Yes Yes

Yes Yes Yes

Duodenal stenosis VSD Bilateral choanal atresia

RM HH

1.98 1.57

TEF TEF

Yes No

No Yes

Yes Yes

Absent right lung GE reflex

VSD, PDA Gastric rupture, IVH, VSD, large ASD Hypoplastic left heart Gastric rupture, VSD, DA

Abbreviations: NA = nonapplicable; VSD = ventricular septal defect; ASD ~ atrial septal defect; PDA = patent ductus arteriosus; NEC = necrotizing enterocolitis; TGA = transposition of great arteries; IVH = intraventricular hemorrhage; PFC = persistent fetal circulation; DA = duodenal atresia, *All patients requiring preop ventilatory support. tEmergency occlusion of fistula attempted via ruptured stomach but failed,

pressure ventilation because of severe RDS. Thirteen patients of this nature have been managed since then. Two patients died before surgical intervention could be undertaken. One of these patients died of hypoplastic left heart syndrome. The other patient weighed only 1.08 kg and was found on postmortem examination to have a ventricular septal defect and duodenal atresia complicated by gastric rupture. Review of the clinical management of this infant indicated that excessive ventilation was responsible for the gastric rupture. A third patient who died in this latter time period also developed preoperative rupture of the stomach while being ventilated for RDS. Following transport to our hospital, she was immediately taken to the operating room for a laparotomy. In spite of attempts to control the air leak with a Foley catheter in the distal esophagus, she died before the fistula could be divided. In the last five years, eight patients with severe RDS underwent division and repair of their TEF as the first step in their managment. Seven of these eight patients had uneventful operations; five of them were so stable that it was elected to proceed with esophageal anastomosis while the other three were staged. One of the eight patients who underwent initial division of the

TEF did develop severe deterioration during the procedure, but she had an absent right lung. Although this child survived operation, significant problems with adequate ventilation of a solitary left lung were encountered during the procedure. The remaining two of the 13 patients managed during this latter period had gastrostomy performed first and both of them had ventilatory problems as soon as the gastrostomy tube was inserted. DISCUSSION

In this group of 22 patients, 11 or 50% survived. If the five patients who died before definitive repair could be attempted are excluded, 11 of 17 or 65% survived. More importantly only four of seven (57%) patients who had gastrostomy performed first survived while seven of ten (70%) who had division of the TEF performed initially survived. The difficulties with intraoperative management of the group who had gastrostomy performed first were even more impressive. Recent reports describe over 90% survival in good risk, Waterston Group A and B undergoing primary repair of EA and TEF. 6-9 On the other hand, the same

396

TEMPLETOET NAL

series continue to show very poor survival rates in Waterston Group C patients, especially those of low birth weight with RDS. Other factors that appear to influence survival adversely in low birth weight infants and TEF are associated life-threatening anomalies, particularly cardiac malformations.8 In recent years, management of the low birth weight infant with EA and TEF has focused on the selection of an initial staging procedure that puts the least stress on the patient, that minimizes secondary complications, and that provides the best opportunity for growth. Classically, this has been primary gastrostomy and staged approaches to the fistula. The occasional but potentially disastrous problem identified in our experience involves that special group of patients with severe RDS who require high pressure ventilatory support within the first few hours of life. Placement of an initial temporizing gastrostomy in such patients resulted in acute ventilatory deterioration in six of the seven patients in whom it was done. Three of them died. Preliminary gastrostomy is still favored by many authors for sick neonates with EA and TEF, but in none of these reports is the specific problem we have identified mentioned. 7'9-tl Two main reasons are given for preliminary gastrostomy, First, it provides drainage of gastric contents so that the lungs are protected from gastric reflux through the fistula. However, regardless of the size of the fistula, review of another group of 62 of our patients with EA and TEF who underwent thoracotomy first did not reveal a single instance of gastric contents in the tracheobronchial tree. Despite the fact that the latter consideration is widely accepted, there is little proof for its validity in

(1) NORMAL LUNGS Spontaneous Ventilation low

;er:::'ve

the literature. Second, when gastric distention is marked, there may be compromise of ventilation and even circulation, l~ The only patients in our experience who incurred marked gastric distention were those whose major ventilatory problem was caused by their underlying lung disease. If gastric distention is a problem, one can pass a catheter through the distal esophagus into the stomach following division of the fistula and easily relieve the distention. Filston 12was the first to emphasize that the problem in patients with a distal TEF and severe RDS was not gastric distention but decreased compliance of the lungs. As shown in Figure 1, effective ventilation of the tracheobronchial tree is very much affected by the compliance of the lungs. In a patient with normal lungs and spontaneous breathing, there is little or no loss of effective ventilation. Some air accumulates in the stomach during the expiratory phase but the intragastric pressures are usually low. Placement of a gastrostomy tube under local anesthesia with spontaneous breathing in such a patient results in no pulmonary compromise. If the same patient with normal lungs is then intubated and given positive low pressure ventilation, some compromise of ventilatory efficiency may develop. Somewhat higher than normal ventilatory volumes and pressures are required because the upper gastrointestinal tract is now functioning in continuity with the tracheobronchial tree. Gastric distention becomes greater. Yet, in the presence of normal lung compliance, the majority of ventilatory exchange still occurs within the lungs and adequate oxygenation can be maintained.

(2) NORMAL LUNGS Positive Pressure Ventilation

~

(3) DECREASED COMPLIANCE Positive Pressure Ventilation

i ~IpP~176 ~

~ ir~ e

Fig 1. Comparison of distending pressures within the tracheobronchial tree and gastrointestinal tract in children with esophageal atresia and distal tracheoesophageal Normal lungs with spontaneous ventilation--no loss of ventilation via normal lungs with low positive pressure ventilation--mild loss of ventilation via poorly compliant lungs with high positive pressure ventilation--marked loss of ventilation via

fistula(1) . TEF.

TEF;(3)

TEF;(2)

MANAGEMENT OF EA AND TEF

397

In contrast, if the patient's lung compliance is poor and pulmonary resistance high, as in the case of a neonate with severe R D S , both the inflating pressure and ventilating volume m a y be increased in order to maintain adequate ventilation. In the presence of a distal T E F , much of this increased ventilatory volume is dissipated as the stomach distends and air enters the small bowel. Two of our patients, in fact, passed flatus in cycle with the ventilator. As a result, the inflating pressure must be increased further to provide adequate oxygenation at the alveolar level. Once again, much of the increased airway pressure is transmitted down the esophagus and into the stomach. Sudden loss of intragastric pressure in this situation m a y disrupt a critical physiologic balance in that most of the inflating pressure is lost. This situation usually occurs the m o m e n t the stomach is opened for placement of a gastrostomy or in the event of rupture of the stomach. Such acute ventilatory deterioration was noted in Filston's patient and in six of our own patients following placement of a gastrostomy. Two of our patients developed a cardiac arrest that could not be corrected until the leak via the fistula was controlled. Proper placement of the tip of the endotracheal tube has been emphasized by many 1~ as the way to prevent such complications in the m a n a g e m e n t of the patient's ventilation. O u r anesthesiologists routinely place the tip of the endotracheal tube just proximal to the carina but distal to the fistula in order to prevent accidental intubation of the fistula or dissipation of ventilation into the stomach. In view of the very short distance between the fistula and the carina, it is not surprising that often this position cannot be strictly maintained. Frequently at thoracotomy, one will note the passage of air down the esophagus and yet, when the lungs are relatively normal, the patient will not be

in distress. In our experience the relative size of the fistula is not a factor in determining whether or not a given patient develops respiratory distress or gastric distension. The key factor is the presence or absence of normal lung compliance. In our experience and in that of others, 14 acute rupture of the stomach is almost always fatal because of the loss of the ability to ventilate the patient effectively. A n emergency laparotomy with occlusion of the leak via the esophagus by means of digital pressure or a Foley catheter m a y permit the patient to be initially resuscitated. In such cases, however, one must proceed with ligation of the fistula if effective ventilation is to be restored. In order to prevent such acute deterioration, Filston 13 suggested that a patient with severe R D S and a distal T E F should undergo bronchoscopy first with placement of a Fogarty catheter to occlude the fistula. The approach, however, requires careful monitoring and periodic adjustments over a period of several days. Moreover, the drawbacks to such an approach are interruption of ventilation while the bronchoscope is being passed, potential technical difficulties in placing the Fogarty catheter in the proximal esophagus, and potential injury to the esophagus by the Fogarty catheter when its presence is required for several days. In conclusion, our experience leads us to suggest that the most expeditious way to inprove ventilatory efficiency in a child with severe R D S and a distal T E F is to expose and ligate the fistula first. In this fashion, R D S may be treated without superimposing the problem of inefficient delivery of ventilation. With this approach, five of our patients improved so dramatically that it was possible to proceed with esophageal anastomosis even though staging was initially contemplated; they all survived.

REFERENCES

1. Brereton R J, Zachary RB, Spitz L: Preventable death in oesophageal atresia. Arch Dis Chid 53:276-283, 1978 2. Koop CE, Hamilton JP: Atresia of the esophagus: Increased survival with staged procedure in the poor-risk infant. Ann Surg 162:389-401, 1965 3. Holder TM, McDonald VG Jr, Woolley MM: The premature or critically ill infant with esophageal atresia: Increased success with a staged approach. J Thorac Cardiovasc Surg 44:344-358, 1962 4. Hays DM, Woolley MM, Snyder WH Jr: Changing techniques in the management of esophageal atresia. Arch Surg 92:611616, 1965 5. Koop CE, Schnaufer L, Broennle AM: Esophageal atresia and tracheoesophageal fistula: Supportive measures that affect survival. Pediatrics 54:558-564, 1974 6. Waterston DJ, Bonham-Carter RE, Aberdeen E: Oesophageal atresia: Tracheo-oesophageal fistula. A study of survival in 218 infants. Lancet 1:819-822, 1962 7. O'Neill JA, Holcomb GW, Neblett WW: Recent experience with esophageal atresia. Ann Surg 195:739 745, 1982

8. Louhimo I, Lindahl H: Esophageal atresia: Primary results of 500 consecutively treated patients. J Pediatr Surg 18:217-229, 1983 9. Abrahamson J, Shandling B: Esophageal atresia in the underweight baby: A challenge. J Pediatr Surg 7:608-613, 1972 10. Dryden GE: Gastric distension during tracheosophageal fistula repair. J Indiana Med Assoc 62:46-47, 1969 11. Salem MR, Wong AY, Lin YH, et al: Prevention of gastric distention during anesthesia for newborns with tracheosophageal fistulas. Anesthesiol 38:82-83, 1973 12. Filston HC, Chitwood WR Jr, Schkolne B, et al: The Fogarty balloon catheter as an aid to management of the infant with esophageal atresia and tracheoesophageal fistula complicated by severe RDS or pneumonia. J Ped Surg 17:149-151, 1982 13. Calverley RK, Johnston AE: The anaesthetic management of tracheoesophageal fistula: A review of ten years' experience. Canad Anaesth Soc J 19:270-282, 1972 14. Jones TB, Kirchner SG, Lee FA, et al: Stomach rupture associated with esophageal atresia, tracheoesophageal fistula, and ventilatory assistance. AJR 135:675-677, 1980