A Bifurcated Tracheal Tube for a Neonate with Tracheoesophageal Fistula

The patient was male, weighed 2,908 g at 41 weeks gestational age, had difficulty feeding, and was transferred to our hospital at 2 days of age. A diagnosis of EA/TEF was made, and cardiac ultrasonography revealed that he had a large patent ductus arteriosus (PDA), atrial septal defect, and mitral and tricuspid valve regurgitation. Because of the severity of his congenital heart disease, radical operation of the EA/TEF was postponed and an emergent gastrostomy was performed using general anesthesia. Intraoperative bronchoscopy revealed a large TEF (> 4 mm in diameter, as large as the patient’s bilateral main bronchi) just proximal to the carina on the dorsal side of the trachea (fig. 1). Proper positioning of the endotracheal tube to avoid ventilation of the fistula 2was impossible because of the short distance from the TEF to the carina. The tube was therefore placed in an ordinary position (above the fistula), and the patient was carefully ventilated with low peak inspiratory pressure in the semi-Fowler position. On completion of surgery, he resumed spontaneous breathing and his trachea was extubated. The next day, the patient exhibited tachypnea, oliguria, and severe metabolic acidosis (pH 7.239, Paco²12.4 mmHg, base excess −21.3mEq/l). Ultrasonography revealed cardiac insufficiency (right ventricular end diastolic diameter, 17.3 mm; left ventricular end diastolic/systolic diameter, 14.2/11.7 mm; ejection fraction, 40%), deterioration of mitral regurgitation, and increased left-to-right shunting through the large PDA, for which an emergent PDA clipping was indicated. After induction of anesthesia, ventilation was difficult because of a massive air leak through the gastrostomy tube. 3–5,7Intragastric pressure, monitored by an aneroid manometer, exhibited inflation of the stomach on each ventilation. We managed to ventilate the lungs by limiting the intragastric pressure to 10–15 cmH²O, which was achieved by intermittent clamping and declamping of the gastrostomy tube.

In preparation for the patient’s EA/TEF repair, we developed a bifurcated tube to deliver air/oxygen directly to the distal airway without ventilating the fistula. The tip of a Portex® endotracheal tube, 3.0 mm ID (SIMS Portex Inc., Keene, NH), was cut to create a bevel facing the right side (fig. 2a). Then the distal end of the tube, approximately 8 mm in length, was incised longitudinally (fig. 2b). When splayed open, it became a bifurcated tip with two half-lumens (fig. 2c and d). The cut surface of the tube was smoothed with fine sandpaper and tetrahydrofuran, 8and the bifurcated segments were left almost closed (fig 2b). We then simulated introducing the tube to sit astride the carina using a neonatal tracheal model, which we constructed by suturing two 4.5-mm-ID tracheal tubes together.

The operation was scheduled for day 8. With the approval of our institute’s ethical committee and the informed consent of the patient’s parents, placement of the bifurcated endotracheal tube was attempted after induction of anesthesia. Because the bifurcated tip made it difficult for the tube to pass through the glottis, we held the midpoint of the bifurcated segments of the tube in mosquito forceps so that the bifurcated tip was closed during passage through the glottis. Once the closed tip passed through the glottis, the forceps were removed and the tube was advanced into the trachea. While maintaining ventilation via  a swivel connector, a fiberscope (model LF-P; Olympus Optical Co. Ltd., Tokyo, Japan) was inserted into the tube and advanced into the left main bronchus, just proximal to the opening of the left upper lobe bronchus (fig. 3a). The tube was then advanced over the fiberscope until the left half of the bifurcated tip was introduced into the left main bronchus. Because the right main bronchus had a smaller branching angle from the trachea than the left main bronchus, the right half of the divided tip tended to enter the right main bronchus, and the bifurcated segments of the tube were forced to splay to sit astride the carina (fig. 3b).

The fiberscope was then withdrawn to the carinal level and proper placement of the tube was confirmed (fig. 4). Ventilation was ascertained by chest movement, auscultation, intragastric pressure monitoring, and arterial blood gas analysis. The increase in intragastric pressure diminished, and the arterial blood gas proved adequate oxygenation without carbon dioxide retention. A chest radiograph also confirmed proper positioning of the tube (fig. 5). On completion of surgery, the bifurcated tube was exchanged for an ordinary 3.0-mm-ID tube. Postoperative bronchoscopy demonstrated no bleeding except for the slight flush of tracheal mucosa around the carina. The patient was transferred to the ICU and was weaned from the ventilator on the second postoperative day. There were no sequelae related to the use of the bifurcated tube.

The most common way to avoid entry of gas into a TEF is to position the endotracheal tube so that the tip is located just proximal to the carina but distal to the fistula. 2This works well if the fistula is located far enough above the carina. However, because the fistula may be located just above the carina, 3it may be difficult to strictly maintain this position. 7A subtle change in the tube position, often caused by surgical manipulation or patient’s movement, could be critical. Therefore, this technique would not be suitable for several days of intensive care for the EA/TEF patient with severe respiratory distress syndrome, pneumonia, or prematurity requiring mechanical ventilation before surgery. 1,4,5,7 

Occlusion of the fistula with a Fogarty embolectomy catheter, either through the trachea (anterograde occlusion 3,4,6) or the gastrostomy site (retrograde occlusion 5), is another strategy that has been reported by several authors. This method can be effective but may not be feasible in every case. 7Disadvantages of the method are as follows:

1. There are potential technical difficulties in placing the catheter into the fistula. In anterograde occlusion, manipulation of a Fogarty catheter into the fistula through the trachea under a thin rigid bronchoscope is technically difficult and requires interruption of ventilation during the procedure. 7Limitation of the size of the rigid bronchoscope could also be a problem. The smallest size available is 4.2 mm OD (DOESEL-HUZLY Pediatric Bronchoscope K10339F; Karl Storz GmbH & Co., Tuttlingen, Germany; size 2.5, 3.5 mm ID, 4.2 mm OD, 20 cm in length), which is equal to the OD of a 3.0-mm-ID Portex® endotracheal tube (4.2 mm). Thus, the method is not applicable to patients too small to accept a 3.0-mm-ID endotracheal tube, 3which is often the case. 1,7Retrograde occlusion is technically complicated and necessitates the placement of a gastrostomy, which is no longer routinely performed in patients with no other complicating factors. 1,3,6,7 

2. The catheter could be dislocated to occlude the trachea, 4,5either by surgical manipulation or the patient’s movement, making it totally impossible to ventilate.

3. The catheter might damage the esophageal mucosa in contact with the balloon by compromising its circulation. 5,7 

Compared with the two strategies mentioned, our bifurcated tube has several advantages:

1. It is easy to prepare and easy to place under fiberoptic guidance.

2. It does not require rigid bronchoscopy and is not limited by the size of the patient’s trachea. By using smaller tracheal tubes, our method can be applied to premature neonates.

3. There may be less risk of accidental displacement with the bifurcated tip sitting astride the carina; thus, it may be useful not only for intraoperative management but also for management of those patients requiring mechanical ventilation before surgery. 1,4,5,7 

4. It does not require interruption of ventilation during the procedure.

5. It may help in the event of tracheomalacia, a common complication in EA/TEF patients. 9 

6. It may also be applicable to a laryngotracheoesophageal cleft, in which a much more complicated use of a bifurcated endotracheal tube with each lumen entering two main bronchi has been reported. 10 

On the other hand, our tube does have its disadvantages. First, the half-round tips may flatten or turn over and occlude the bronchi, especially when the tube is inserted too far. If the tube is too small, the two tips might fold backward as the tube passes down the trachea. However, using an adequately sized tube and maintaining the correct position of the tube can avoid these complications.

Second, the bifurcated tube might compromise the ventilation of the right upper lobe, especially when the right upper lobe bronchus is very close to the carina or a tracheal bronchus exists. As the length of the right main bronchus (the length of the right main bronchus to the origin of the right upper lobe bronchus) in neonates is reported to be 9.92 ± 1.67 mm (mean ± SD), 11we consider that the right side of the bifurcation should be shorter than 6.48 mm (mean − 2 SD value) at the lateral aspect of the tube to avoid occlusion of the right upper lobe bronchus. However, when the bifurcation becomes shorter, placing the bifurcated tip to sit astride the carina and avoiding dislocation of the tube would be more difficult. In the trials using the neonatal tracheal model, we found that the bifurcated tip could not easily be placed to sit astride the carina when the longitudinal incision was shorter than 8 mm. Thus, we decided the incision should be 8 mm long, which resulted in a 5.5-mm distance from the carina to the distal end of the right bifurcated segment of the tube at the lateral aspect of the tube (fig. 2b). Assuming that an abnormal origin of the right upper lobe bronchus did not exist, there was little risk of occluding the right upper lobe bronchus because the distance was shorter than 6.48 mm.

Finally, in rare cases, the bifurcated tube is ineffective when a TEF exists at the bronchial level.

Another controversial point of managing this case was the associated major congenital heart disease. EA/TEF patients often have complicated congenital heart disease, which can alter the outcome and management of TEF. However, heart disease is not necessarily an impediment to primary repair of the EA/TEF. 3,12In our case, the patient’s PDA was so large that circulatory deterioration was anticipated to occur during primary repair, which is performed in the left decubitus position. Therefore, we decided to perform a gastrostomy and to delay the surgical repair of the EA/TEF. As we had anticipated, the patient deteriorated soon and subjected to emergent PDA ligation.