- Email: [email protected]
Repair of congenital tracheal stenosis
Repair of Congenital Tracheal Stenosis Carl L. Backer, Constantine Mavroudis, and Lauren D. Holinger We have used six different techniques in 61 operations performed at Children’s Memorial Hospital (Chicago, IL) between 1982 and 2001 on 54 infants with complete tracheal rings and primary or recurrent tracheal stenosis. Short-term and long-term outcomes are reviewed for all techniques including pericardial tracheoplasty, tracheal autograft, tracheal resection, cartilage tracheoplasty, slide tracheoplasty, and aortic homograft patch tracheoplasty. Patients’ ages ranged from 7 days to 72 months; mean age at operation was 6 months. There were 33 boys and 21 girls. Associated pulmonary artery sling was present in 19 of 54 patients (35%). Significant associated intracardiac anomaly was present in 13 of 54 patients (24%). The number of complete tracheal rings involved ranged from two to 18 rings (mean, 14 ⴞ 5). All procedures were performed with the use of cardiopulmonary bypass. Simultaneous repair of pulmonary artery sling and cardiac abnormalities was undertaken. There were three early deaths, two after pericardial tracheoplasty and one after tracheal autograft. There were eight late deaths, five after pericardial tracheoplasty, one after tracheal autograft, one after slide tracheoplasty, and one after tracheal resection. Median length of hospital stay was 60 days for pericardial tracheoplasty, 28 days for tracheal autograft, 14 days for tracheal resection, and 18 days for the slide tracheoplasty. Follow-up is complete in all patients. Tracheal autograft is currently our procedure of choice for patients with long-segment tracheal stenosis because of its use of all-autologous material, technical ease of performance, already-present epithelial lining of the autograft, intrinsic maintenance of the cartilage contour, potential for growth, and ready availability. We limit the use of tracheal resections to patients with eight or less rings of tracheal stenosis. Copyright © 2002 by W.B. Saunders Company Key words: Tracheal stenosis, tracheal autograft, tracheoplasty, pulmonary artery sling, tracheal rings.
T
he most frequent congenital tracheal anomaly requiring operative intervention is tracheal stenosis secondary to complete tracheal rings. This occurs when there is absence of the posterior membranous trachea and the cartilage of the trachea is circumferential (Fig 1). The tracheal orifice in these patients may be exceedingly small and cause presentation with severe respiratory distress within the first several weeks of life. Tracheal stenosis has been categorized by Cantrell and Guild1 into three major subsets; segmental (localized) stenosis, funnel-like stenosis, and generalized hypoplasia (long-segment stenosis). In the late 1970s and early 1980s, most case reports regarding this lesion reported the
From the Division of Cardiovascular-Thoracic Surgery, the Division of Pediatric Otolaryngology, and the Department of Communicative Disorders, Children’s Memorial Hospital, and the Department of Surgery, Northwestern University Medical School, Chicago, IL. Address reprint requests to Carl L. Backer, MD, Division of Cardiovascular-Thoracic Surgery-M/C#22, Children’s Memorial Hospital, 2300 Children’s Plaza, Chicago, IL 60614. Copyright © 2002 by W.B. Saunders Company 1092-9126/02/0501-0006$35.00/0 doi: 10.1053/pcsu.2002.29718
diagnosis but had a fatal outcome.2,3 Benjamin et al’s4 review of the medical management of these patients reported a 43% mortality rate. Many of these patients have associated anomalies such as pulmonary artery sling and severe intracardiac congenital anomalies that contribute to the morbidity and mortality of this patient population. The first successful procedure for tracheal stenosis secondary to complete tracheal rings was reported by Kimura et al5 in 1982. They used a rib cartilage graft to perform an anterior tracheoplasty. This was followed in 1984 by the series reported by Idriss et al6 who performed tracheoplasty with an autologous pericardial patch for extensive tracheal stenosis. Jonas et al7 reported tracheal repair by tracheal resection in 1989. The slide tracheoplasty was also reported in that year by Tsang et al.8 Elliott et al9 reported the use of cadaveric homograft material for tracheal reconstruction in 1996. We reported the use of a free tracheal autograft in 1998.10 At Children’s Memorial Hospital (Chicago, IL), the surgical advance of pericardial tracheoplasty devised by Dr Idriss and the strong support from our otolaryngology division established our
Pediatric Cardiac Surgery Annual of the Seminars in Thoracic and Cardiovascular Surgery, Vol 5, 2002: pp 173-186
173
174
Backer, Mavroudis, and Holinger
service as a center of referrals for tracheal stenosis. This has allowed us to develop a comprehensive program combining the respective skills of cardiothoracic surgery and otolaryngology. Over the past 19 years we have used six different techniques for infants with primary or recurrent tracheal stenosis. These techniques include pericardial tracheoplasty, tracheal resection, cartilage tracheoplasty, tracheal autograft, slide tracheoplasty, and aortic homograft patch tracheoplasty. Our techniques have evolved over time as we have gained experience with the advantages and disadvantages of the different techniques. This review will examine the short-term and long-term outcomes of our patients with congenital tracheal stenosis and review the world literature with regard to tracheal surgery in infants and children.
Patients and Methods
Figure 1. Illustration of a patient with complete tracheal rings affecting approximately two thirds of the tracheal length. The upper cutaway shows a normal tracheal ring with the anterior cartilage and relatively flat posterior membranous trachea. In the midportion, a complete tracheal ring is shown which is a circumferential cartilage ring with no membranous trachea. The lumen at the site of the complete tracheal ring can be quite small. (From Backer, CL, Mavroudis C, Gerber ME, et al: Tracheal surgery in children: An 18-year review of four techniques. Eur J. Cardiothorac Surg, vol 19, pp 777-784, 2001. Reprinted with permission.)
Between 1982 and 2001, 54 consecutive children with complete tracheal rings have undergone surgical repair at Children’s Memorial Hospital (see Table 1). The objective of this chapter is to review the short-term and long-term outcomes of the six different techniques used in these patients. The patients’ ages ranged from 7 days to 72 months with a mean age of 6 months. There were 33 boys and 21 girls. Nineteen of these patients had an associated pulmonary artery sling; thirteen had a significant associated intracardiac anomaly. All procedures were performed with the use of cardiopulmonary bypass. Simultaneous repair of pulmonary artery sling and cardiac abnormalities was undertaken. Symptoms at the time of presentation included respiratory distress, apnea, cyanosis, and stridor. Often these
Table 1. Techniques Used for Tracheoplasty at Children’s Memorial Hospital (Chicago, IL) Technique
Primary Secondary (No. of Pts) (No. of Pts)
Pericardial patch Tracheal autograft Tracheal resection Slide tracheoplasty Cartilage tracheoplasty Aortic homograft patch
28 15 9 2 0 0
2
4 1
Total
54
7
Repair of Congenital Tracheal Stenosis
patients presented as “difficult intubations” at the time of attempted intubation for either other surgical procedures or for respiratory distress. Many of these patients were referred with an endotracheal tube quite small for the patient’s age placed just through the vocal cords. Several of these patients had extremely high CO2 levels at the time of presentation. All patients had rigid bronchoscopy, either a day or two before the operative intervention or at the same time. All recent patients have had transthoracic echocardiography to rule out a pulmonary artery sling.11 Children who required only an isolated tracheal operation were placed on cardiopulmonary bypass with an aortic cannula and a single atrial cannula. These patients were cooled to 32°C and remained in normal sinus rhythm throughout the procedure. If the patient had an associated pulmonary artery sling, the repair of the sling was performed while on cardiopulmonary bypass. This was performed before the tracheal repair to prevent contamination of the vascular anastomosis by the open tracheal lumen. The technique of pulmonary artery sling repair was to transect the left pulmonary artery from its origin on the right pulmonary artery, dissect the left pulmonary artery free of the posterior mediastinal attachments behind the trachea and anterior to the esophagus, and move the left pulmonary artery anterior to the trachea. The left pulmonary artery was then reimplanted into the main pulmonary artery at a site that approximated where the normal left pulmonary artery would have originated.12 This was typically either adjacent to or at the site of a ligated ligamentum or patent ductus arteriosus. The left pulmonary artery anastomosis was commonly performed with interrupted PDS II (polydioxanone; Ethicon, Inc, Somerville, NJ) or Prolene (polypropylene; Ethicon, Inc) suture. If there was an associated intracardiac lesion that required repair, bicaval venous cannulation was used. The intracardiac lesion was repaired during a period of aortic cross clamp time before the tracheal repair, again, attempting to avoid contamination of any intracardiac material such as Dacron or Gore-Tex (W.L. Gore & Assoc, Flagstaff, AZ) by the open tracheal lumen.
Pericardial Patch Tracheoplasty This operation was our procedure of choice between 1982 and 1995. Immediately following the
175
median sternotomy, an autologous pericardial patch is harvested. This patch extends from the diaphragmatic surface to the reflection of the pericardium at the great vessels, and is typically 5 to 6 cm in length and 2 to 3 cm in width. The anterior surface of the trachea is dissected extensively from the cricoid to the carina. After cardiopulmonary bypass has been initiated, a number 11 blade is used to incise the trachea anteriorly throughout the extent of the complete tracheal rings. In the early part of our experience, we performed this under bronchoscopic guidance. Most recently we used direct visualization of the complete tracheal rings from anteriorly to guide the extent of this incision. The pericardial patch is brought onto the field and sutured in place using interrupted 6-0 Vicryl sutures (polyglactin 910; Ethicon, Inc) (Fig 2). By starting the suturing at the most inferior aspect of the incision adjacent to the carina, the accumulation of blood and secretions in the tracheal lumen is minimized. We used a running suture technique for only one patient in this series. That patient developed a patch dehiscence and lethal pseudomonas mediastinitis. Following placement of the patch, rigid bronchoscopy is performed as cardiopulmonary bypass is continued. This confirms relief of stenosis and is used to suction the tracheobronchial tree clear of blood and secretions. The endotracheal tube is then reintroduced so that the tip is just above the carina. The endotracheal tube effectively stents the patch open. Two small hemoclips are placed at the superior and inferior aspects of the patch. This helps to radiographically identify the extent to the patch and its location vis-a`-vis the endotracheal tube in the postoperative period. The mediastinum is filled with saline and the patient is ventilated to a peak airway pressure of 40 cm H2O to assess for any leaks in the patch. Leaks are sutured closed with interrupted Vicryl suture. The patient is then ventilated, warmed, and weaned from cardiopulmonary bypass. In the current era, we perform modified ultrafiltration for 10 minutes. The patch is sealed with Tisseel glue (Baxter Healthcare Corp, Glendale, CA). Standard chest closure is performed. The patient is kept paralyzed, intubated, and ventilated for 1 week. At that time, bronchoscopic assessment is performed to look for healing of the patch, sites of residual stenosis, and to remove granulation
176
Backer, Mavroudis, and Holinger
ing of the patch and to remove granulation tissue as required.
Tracheal Autograft
Figure 2. Illustration of a completed pericardial patch tracheoplasty. The pericardial patch has been sutured in the anterior tracheal opening. The pericardium is anchored with interrupted Vicryl sutures. In cross-section (inset), the tracheal lumen is shown now to be normal in size. (From Backer CL, Mavroudis C, Gerber ME, et al: Tracheal surgery in children: An 18-year review of four techniques. Eur J Cardiothorac Surg, vol 19, pp 777-784, 2001. Reprinted with permission.) tissue. Over the next 1 to 2 weeks the patient is weaned from the ventilator and is usually extubated 2 to 3 weeks after the procedure. Serial bronchoscopies are performed to assess the heal-
Tracheal autograft has been our procedure of choice since 1996. The tracheal autograft technique is performed with a median sternotomy approach and the use of cardiopulmonary bypass. The entire trachea is extensively mobilized circumferentially. This allows eventual resection of a significant portion of the midtrachea. The isthmus of the thyroid gland is divided. Dissection is carried below the carina well onto the right and left mainstem bronchi. Following dissection, the anterior trachea is incised in the area of the complete tracheal rings with a number 11 blade (Fig 3, left). With the tracheal lumen open through the rings, an assessment is made as to how much trachea can be resected without causing excessive tension at the eventual posterior anastomosis (see Figs 3, center and right). The tracheal autograft is harvested by excising the midportion of the open trachea. The autograft is usually 2 cm in length. The two remaining ends of the trachea are anastomosed posteriorly with multiple interrupted 6-0 PDS sutures (Fig 4, left). The sutures are placed to keep the sutures and knots out of the tracheal lumen. This helps to diminish the incidence of granulation tissue formation at the suture lines. The tracheal autograft that was previously harvested is used to patch the remaining anterior opening in the trachea. The corners of the autograft are trimmed (Fig 4, center). Sometimes the autograft has to be made smaller by cutting 1 to 2 mm along the length of each side of the autograft. The autograft is anchored in place with interrupted 6-0 PDS sutures (Fig 4, right). If the autograft is not long enough to augment the existing opening in the trachea, a composite with pericardium is made and the inferior portion of the anterior tracheal opening is patched with the autograft. The autograft is trimmed in a different fashion, not cutting the corners of the superior aspect of the autograft. The remaining upper opening in the trachea is then patched with fresh autologous pericardium, again anchored with 6-0 PDS sutures (Fig 5). This, of necessity, results in an anastomosis between pericardium and the tracheal autograft. Again, the suture line is tested for air leaks by ventilating the patient to 40 cm H2O pressure. Tisseel glue is again used to seal
Repair of Congenital Tracheal Stenosis
177
Figure 3. Tracheal autograft. An anterior longitudinal incision is performed through the complete extent of the tracheal rings (left). The midportion of the trachea is excised to be used as the tracheal autograft (center). The two remaining orifices of the trachea are reapproximated posteriorly (right). (Reprinted with permission.10)
the suture line. Two or three hemoclips are placed to identify the upper and lower aspect of the autograft or the upper and lower aspect sequentially of the pericardium and the autograft. The strategy for endotracheal tube placement depends on the type of autograft. If there is no pericardial augmentation, the endotracheal tube is kept above the area of the autograft. If the pericardium is used, the midportion of the tube is placed at the midpoint of the tracheal autograft. The tube will then stent the pericardium open. In our series, no patient had a neck brace or a chin stitch. The patient is kept paralyzed and ventilated for 3 to 5 days, at which time follow-up bronchoscopy is performed. If the tube is above the area of the anastomoses, this can be done with flexible bronchoscopy. If the tracheal lumen is patent with minimal granulation tissue and the child is stable from a respiratory and hemodynamic standpoint, they are weaned from the ven-
tilator over the next 3 to 5 days. Bronchoscopy is again performed just before extubation and then electively just before hospital discharge and at 3- and 6-months postdischarge. Some patients who have formation of granulation tissue or residual stenosis may require bronchoscopy more frequently.
Tracheal Resection Tracheal resection is applicable only to patients with short-segment (localized) tracheal stenosis. In our series, all tracheal resections have been performed with the use of cardiopulmonary bypass. The trachea is mobilized in a fashion similar to that described for the tracheal autograft. If there is any question about the site of the tracheal stenosis, rigid bronchoscopy is performed during cardiopulmonary bypass and a 25-gauge needle is used to penetrate the trachea. The needle is identified endoscopically and used to
178
Backer, Mavroudis, and Holinger
Figure 4. Tracheal autograft. The posterior anastomosis is performed with interrupted PDS sutures (left). The autograft is prepared by trimming the corners of the autograft (center). The autograft is sutured in place to cover the remaining opening in the anterior trachea (right). (Reprinted with permission.10)
guide the surgeon as to the length of the tracheal resection. In most cases, an anterior tracheotomy is made in a longitudinal fashion with a number 11 blade through the area of the stenotic rings. This confirms the location of the rings and allows one to convert to the tracheal autograft technique should the length of tracheal stenosis be longer than expected from the preoperative evaluation. In most cases a bilateral hilar release, freeing of the pulmonary artery from the pericardium, and in some instances a hyoid release are used to provide enough mobility to perform a successful end-to-end anastomosis. This anastomosis is performed with interrupted 6-0 PDS suture (Fig 6). In most cases, the endotracheal tube is positioned above the anastomosis; however, in some cases the tube was passed through the anastomosis. These patients were kept paralyzed for a much shorter period of time, 24 to 48 hours, and then were usually weaned from the ventilator and extubated. Most patients had one bronchoscopy just before extubation and one just before hospital discharge.
Slide Tracheoplasty We used slide tracheoplasty for two patients in 1996. Again, the patient is approached through a median sternotomy with the use of cardiopulmonary bypass. Slide tracheoplasty requires mobilization of the trachea in its entirety, similar to the autograft and tracheal resection techniques. The exact midportion of the stenosis must be identified because the trachea must be transected at this point (Fig 7). We have used bronchoscopic identification to guide us to the midpoint of the tracheal stenosis, despite what the external appearance of the trachea might be. The superior portion of the trachea is then incised posteriorly through the extent of the complete tracheal rings. In a similar fashion, the inferior aspect of the trachea is incised anteriorly, again through the extent of the complete tracheal rings. At both ends of the tracheal transection the corners are trimmed so that the leading edge will fit into the V-portion of the other component (Fig 8). The two components are slid together and anastomosed with interrupted 6-0 PDS sutures (Fig 9).
Repair of Congenital Tracheal Stenosis
179
Figure 5. Composite tracheal autograft. (A) The posterior anastomosis is performed. (B) The autograft is trimmed, only cutting the inferior corners of the autograft. A portion of pericardium is harvested and tailored. (C) The autograft is sutured in place anteriorly adjacent to the carina. The pericardial patch is inserted superiorly to complete the repair. (From Backer CL, Mavroudis C, Gerber ME, et al: Tracheal surgery in children: An 18-year review of four techniques. Eur J Cardiothorac Surg, vol 19, pp 777-784, 2001. Reprinted with permission.)
The completed anastomosis results in a trachea that is half as long and has four times the luminal diameter (Fig 10). The endotracheal tube is positioned in the midportion of the trachea. We and others8 have managed these patients postoperatively in a fashion very similar to those patients who have undergone a tracheal resection.
Cartilage Tracheoplasty Cartilage tracheoplasty has been described by several surgeons as a primary strategy for the patient with complete tracheal rings. In our series, however, we have used it only as a secondary operation after pericardial tracheoplasty. These patients typically have had distal tracheal stenosis and were patients who presented both with complete tracheal rings and a pulmonary artery sling. In these patients the pulmonary artery
sling affected the distal trachea, causing severe distal tracheal stenosis and often right mainstem bronchial stenosis. These patients were all operated on through a repeat sternotomy with the use of cardiopulmonary bypass. The cartilage graft was harvested from the fourth or fifth anterior cartilage on the right side. This cartilage is excised in toto and then cut longitudinally in half. Our otolaryngology colleagues, who commonly use these grafts for laryngotracheal reconstruction in the neck, then assist with the cartilage shaping. This results in a fairly thin cartilage with a step at the edge for the insertion of the cartilage into the tracheal lumen. The distal trachea is incised under bronchoscopic guidance through the area of stenosis. This was typically from the distal trachea out onto the right main-
180
Backer, Mavroudis, and Holinger
autograft became infected with pseudomonas, and during the time of an emergent bronchoscopy four days postoperatively, the autograft was disrupted. The child had emergency resternotomy, placement on cardiopulmonary bypass, and removal of the autograft. A cryopreserved aortic homograft was then used for the tracheal reconstruction. The aortic homograft was cut in a fashion similar to the patch we previously had used as a pericardial patch. After inserting the homograft in place with interrupted PDS sutures, the pectoralis major muscle was brought through an intercostal window to treat the child’s mediastinitis.13 The child had a very stormy postoperative course requiring balloon expandable Palmaz tracheal stents (Cordis Corp, Miami, FL)
Figure 6. Tracheal resection. The localized segment of tracheal stenosis is excised. The trachea is extensively mobilized, specifically mobilized in the right and left mainstem bronchus and carina. This allows the trachea to be brought together without tension for an end-toend anastomosis using interrupted PDS sutures.
stem bronchus. The cartilage is then fixed in place with interrupted Vicryl or PDS sutures (Fig 11). The patients were managed postoperatively with the endotracheal tube in the midportion of the trachea, similar to a patient after tracheal resection.
Aortic Homograft One patient in our series, following tracheal autograft, had dehiscence of the autograft patch. This was a 2-month-old boy who underwent complete repair of pulmonary artery sling, tetralogy of Fallot, and had a tracheal autograft. The child had a tracheal right upper lobe bronchus. This is the only child in the series in whom the autograft was divided in half longitudinally. These halves were used separately to patch the tracheal opening in an effort not to use pericardium. The
Figure 7. Slide tracheoplasty. The trachea is transected at the midpoint of the long-segment congenital tracheal stenosis. The superior portion of the trachea is incised posteriorly, the inferior aspect of the trachea is incised anteriorly. (Reprinted with permission.16)
Repair of Congenital Tracheal Stenosis
181
Figure 8. Slide tracheoplasty. The upper trachea has been opened posteriorly, the lower trachea has been opened anteriorly. The corners of the transected trachea are trimmed so that the leading edge will fit into the Vportion of the other component. (Reprinted with permission.16)
at 4 weeks and 3 months postoperatively. He had a tracheostomy at 7 months postoperatively. His hospital course was nearly 11 months and then he was discharged home. During bronchoscopy at the referring institution 6 months later the trachea was perforated and the patient died.
Results In this series of 61 operations on 54 children, there were three early deaths; two after pericardial tracheoplasty and one after tracheal autograft. There were eight late deaths, five after pericardial tracheoplasty, one after tracheal autograft, one after slide tracheoplasty, and one after tracheal resection. Nineteen patients had a pulmonary artery sling and 13 patients had a significant intracardiac anomaly. The associated intracardiac anomalies are shown in Table 2. The number of complete tracheal rings involved ranged from two to 18 rings (mean 14 ⫾ 5). Nine patients had a tracheal right upper lobe bronchus. Eight patients had a patent ductus arteriosus, three had an absent right lung, two an absent left lung, and two a severely hypoplastic right lung. In this series there was no complication specifically related to the use of cardiopulmonary bypass and no patient required a reop-
Figure 9. Slide tracheoplasty. This illustrates the beginning of the long anastomotic suture line that is performed with interrupted 6-0 PDS suture while the patient is on cardiopulmonary bypass. (Reprinted with permission.16)
182
Backer, Mavroudis, and Holinger
Table 2. Associated Significant Cardiac Anomalies Anomaly
Number
Tetralogy of Fallot Ventricular septal defect Complete atrioventricular septal defect Pulmonary atresia-Ventricular septal defect Double outlet right ventricle Atrial septal defect
4 4 2 1 1 1
Total
13
tracheoplasty. Follow-up is complete in all patients. Figure 10. Anterior (left) and lateral views (right) of the anastomosed trachea. The tracheal length has been reduced by almost half and the internal luminal diameter has been increased by four times. The cross-sectional appearance of the trachea is shown in the inset. (Reprinted with permission.16) eration for bleeding. The median length of hospital stay was 60 days for pericardial tracheoplasty, 28 days for tracheal autograft, 14 days for tracheal resection, and 18 days for the slide
Pericardial Tracheoplasty Mean age of the patients in this group was 6.5 ⫾ 5.5 months; median age was 6 months. This operation was performed between 1982 and 1995 as a primary procedure on 28 patients and as a secondary procedure on two patients. There were two early deaths. One child died 2 weeks postoperatively after developing pseudomonas mediastinitis from an air leak from the patch. This child was the only patient in this series where a running suture technique was used for the patch. The second death occurred in a 2-week-old infant
Figure 11. Cartilage tracheoplasty. The distal tracheobronchial stenosis has been opened by cutting through the pericardial patch and out onto the right mainstem bronchus. The cartilage craft is carved, inserted into the tracheal opening with the perichondrium facing the lumen, and anchored with interrupted PDS suture. (Reprinted with permission.15)
Repair of Congenital Tracheal Stenosis
with an absent right lung. This child had severe pulmonary hypertension following the procedure and required extracorporeal membrane oxygenation support in the operating room. The child died of an intraventricular hemorrhage 1 week after the operation. There were five late deaths after pericardial tracheoplasty. Two patients died of complications of perforation of the trachea at the time of bronchoscopy at the referring institution (7 and 12 months postoperatively). One child died the day after a tracheostomy 5 months postoperatively. One child died 15 months postoperatively of pulmonary hypertension; that child had also had atrioventricular septal defect repair at the time of tracheoplasty. The final late death was in a child with multiple anomalies who had colonic perforation 22 months following pericardial tracheoplasty. Six of the patients in the pericardial tracheoplasty series required reoperation for recurrent tracheal stenosis. Two patients had a second pericardial patch inserted, four patients had placement of a rib cartilage graft between 2 and 6 months after the pericardial patch. Three patients required five balloon expandable Palmaz stents.14 Six patients required a tracheostomy. In this subgroup, significant risk factors for reoperation included younger age at initial repair, associated pulmonary artery sling, and tracheal right upper lobe bronchus.15
Tracheal Autograft The tracheal autograft operation was first performed in January 1996 and has now been used in 15 consecutive patients. Mean age of these patients was 6 ⫾ 6 months; median age was 4.5 months. Ten of the 15 patients were distant referrals. The mean number of complete tracheal rings was 15 rings. The length of the autograft ranged from 1.3 to 2.5 cm (mean, 2 cm). This was typically six to eight complete tracheal rings. Five of the 15 patients required a pericardial augmentation of the superior aspect of the reconstruction. In those patients, the mean length of the pericardial patch was 2.5 cm. There was one early death (7% early mortality) and there has been one late death (7% late mortality). The one early death occurred 26 days postoperatively in a 6-month-old infant who had previously had two sternotomies for palliation of tetralogy of Fallot. The child had multiple anomalies and required extracorporeal membrane oxygenation support
183
postoperatively because of low cardiac output. He died of sepsis and anasarca. Autopsy showed multiple myocardial infarctions and necrotizing pneumonia, but a widely patent healed tracheal autograft. The late death occurred as described previously in the patient who required an aortic homograft rescue procedure for a failed tracheal autograft. One other patient had significant postoperative complications consisting of recurrent tracheal stenosis related to the pericardial patch portion of the repair. That patient required a Palmaz stent at 3 months postoperatively and a tracheostomy at 4 months postoperatively. That patient had a prolonged (5 months) hospital course but is now 3 years postoperatively and doing well. Three patients required temporary tracheostomies. The other patients in the series are currently asymptomatic from a respiratory standpoint.
Tracheal Resection Nine patients have undergone tracheal resection for shorter segments of tracheal stenosis between 1990 and 2000. The mean age of these patients was 14 months; median age was 4 months. The number of complete tracheal rings excised ranged from two to eight rings with a mean of five rings. There was one late death following tracheal resection. None of these patients required a tracheostomy. Mean hospital stay was 14 days. The one late death occurred 2.5 months after repair of pulmonary artery sling and resection of complete tracheal rings. This child had multiple anomalies including an imperforate anus (status post colostomy), intestinal malrotation, and biliary atresia. The child had end-stage liver failure and was on the waiting list for a liver transplant.
Slide Tracheoplasty Two slide tracheoplasties were performed at our institution, both in 1996.16 One child was 3 months old, had 18 complete tracheal rings, and had simultaneous repair of double outlet right ventricle. Her hospital course was smooth and she was discharged 18 days postoperatively. The other child had pulmonary atresia and ventricular septal defect with 12 complete tracheal rings. She underwent simultaneous modified BlalockTaussig shunt placement and slide tracheoplasty. She developed recurrent tracheal stenosis and granulation tissue. The stenosis was secondary to a “figure-of-eight” configuration of the tracheal
184
Backer, Mavroudis, and Holinger
cross-section. The tracheal rings appeared to spring back to their original cross-sectional area. That child required a Palmaz tracheal stent and tracheostomy. She was hospitalized for 5 months and was unable to be weaned from the ventilator. She died of sepsis from aortic valve endocarditis.
Cartilage Tracheoplasty Four patients underwent reoperation for tracheal stenosis after pericardial tracheoplasty with a cartilage graft.15 One of these patients had a tracheostomy before and after the revision. Two other patients required tracheostomy after the revision. The number of days to hospital discharge ranged from 20 to 105, with a mean of 64 days. The first patient is currently asymptomatic and has a healed patent airway. The second patient died when the trachea was perforated during bronchoscopy at the referring institution 1 year postoperatively. The third patient required two Palmaz stents and a tracheostomy but is currently asymptomatic. The fourth patient required two stents, one in the trachea and one in the left main bronchus, and is currently asymptomatic with a tracheostomy.
Discussion Infants with tracheal stenosis secondary to complete tracheal rings present a significant challenge to the surgeon. These children often present with life-threatening respiratory distress and are admitted to the hospital in extremis. Management of these patients requires close collaboration between the pediatric cardiac surgeon and the pediatric otolaryngologist. The experience of our institution has been an evolution in the surgical procedures for these patients as we have gained experience with the different techniques and their attendant complications. Initially, pericardial tracheoplasty was the operation of choice at our institution and was used successfully for almost 15 years. Pericardial tracheoplasty had a low operative mortality, but because of difficulties with granulation tissue, patch tracheomalacia, and the attendant prolonged hospital course, we devised another technique that has become our standard procedure for these patients. In 1996 we attempted two slide tracheoplasties with mixed results. The first patient had a smooth postoperative course and recovery. The
second patient, however, had significant complications from granulation tissue and recurrent stenosis. In addition, there are some technical aspects of slide tracheoplasty that are difficult to deal with. The trachea must be transected at the exact midpoint of the stenosis otherwise the resultant tracheal openings will not correspond in length. When performing the slide tracheoplasty the upper trachea has to be lifted anteriorly for an initial series of posterior sutures that are difficult to see and to tie. Air leaks along this posterior suture line are particularly difficult to deal with.17 In late 1995 we became aware of the efforts of several groups in Europe who used tracheal homografts for rescue procedures in patients who had failure of a previous operation for congenital tracheal stenosis. Elliott et al9 reported the use of tracheal homograft for these critically ill patients. At this same time, we operated on a patient with an extremely long trachea and decided to resect a number of the tracheal rings, performing an end-to-end anastomosis posteriorly before the pericardial tracheoplasty. The removed tracheal rings were sent to pathology. After reading the report by Elliott et al, we were intrigued by the fact that the piece of trachea that had been discarded looked exactly like the tracheal homograft that they were using. Because of this, for the next patient in our series ( January 1996) who underwent repair of tracheal stenosis, we used the midportion of the trachea as a distal autograft. That child had a very diminutive trachea, pulmonary artery sling, and a ventricular septal defect. The postoperative course was very good and the child was discharged from the hospital 20 days postoperatively. Because of this initial success we had great enthusiasm for this technique and have now used it in the next 14 patients. Initially, we had fears that the autograft might necrose and that in patients who required pericardial augmentation there might be dehiscence of the segment between the autograft and the pericardium. No patient in our series has had either of these complications. Only one patient has had autograft dehiscence and we feel this was caused by the technical issues of making the lumen too small combined with a pseudomonas infection. In no patient was there nonhealing of the junction between the autograft and the pericardium. The autograft usually appears pale but pink, even
Repair of Congenital Tracheal Stenosis
during the initial bronchoscopy. By 5 to 7 days postoperatively, it almost appears to have normal pink mucosal color. Several of the patients have had a very dusky appearing autograft for the first one or two bronchoscopies, but again, none have undergone autograft lysis. In our laboratory, Dr Dodge Khatami performed a variation of the autograft operation on 30 rabbits.18 We had no autograft dehiscence or ischemia in these rabbits. In that model we showed that the topical application of vascular endothelial growth factor enhanced the healing of the autograft. The use of vascular endothelial growth factor resulted in reduced luminal stenosis, reduced mucosal fibrosis, and reduced inflammatory infiltrate. The tracheal architecture was better preserved in the vascular endothelial growth factor-treated animals, and they had a greater microvascular density of the healing tracheal wall. Currently, the tracheal autograft is our procedure of choice for patients with long-segment tracheal stenosis. The advantages of the tracheal autograft include use of all autologous material for the repair, technical ease of performance with all-anterior suturing, already-present epithelial lining of the autograft, intrinsic maintenance of the cartilage contour, potential for growth, and ready availability because the trachea in these infants often appears excessively long. The technique of tracheal resection works extremely well with infants with shorter segment tracheal stenosis. Jonas et al7 reported success with this technique in two patients in 1989. Heinemann et al19 reported success in nine of 10 infants who had resection of tracheal stenosis using extracorporeal circulation. Cotter et al20 recently reported on six patients who all underwent resection with end-to-end anastomosis with excellent results. In our current practice we limit the use of tracheal resections to patients with eight or less rings of tracheal stenosis. Extensive mobilization of the trachea allows the tracheal anastomosis to be performed without tension. These patients have short time to extubation and a shorter hospital stay than many of the other techniques in our series. In the world literature the operative mortality for tracheal resection in infants is approximately 8%.7,19,20 The use of cartilage grafts has been reported by several surgeons to be a successful tech-
185
nique.5,21,22 However, other surgeons have reported a rather high mortality for these patients.23 Our personal opinion is that cartilage is not as easily molded to the trachea as is pericardium or autograft for long-segment stenosis. It tends to maintain its natural torque that does not necessarily conform to the longitudinal tracheal opening. Because of the rigidity of the cartilage it is sometimes difficult to obtain an airtight seal. We have had the experience where the cartilage seems to “fall” into the trachea and actually obstructs the tracheal lumen. The use of aortic homograft has been successful both for primary repair of long-segment congenital tracheal stenosis and for repair of the trachea after extensive injury from trauma.24,25 Aortic homograft is easily available in most pediatric cardiac surgical centers, thus use of the aortic homograft should be a part of any surgeon’s potential armamentarium for these patients. However, because of problems similar to those associated with the use of pericardium, ie, patch tracheomalacia, granulation tissue, and the use of foreign material, we tend to avoid aortic homograft except for emergency situations. Although Elliott and Jacobs have reported good results with tracheal homograft, they have continued to limit this to patients undergoing rescue operation after failed prior procedures.9,26 The tracheal homograft has been reported to result in erosion into surrounding structures and probably should still be considered only as a rescue operation. The infant with tracheal stenosis secondary to complete tracheal rings is a challenge to the managing surgeon. These infants require close cooperation between pediatric cardiac surgeons and pediatric otolaryngologists. Diagnosis of these patients is achieved by rigid bronchoscopy. We have used echocardiography to rule out associated pulmonary artery sling (present in nearly one third of these patients). We recommend repair of the tracheal stenosis through a median sternotomy approach with the use of cardiopulmonary bypass. We currently recommend tracheal resection and end-to-end anastomosis for those patients who have eight complete tracheal rings or less. For patients with long-segment congenital tracheal stenosis, we recommend repair with a tracheal autograft technique. This can be performed with or without pericardial augmen-
186
Backer, Mavroudis, and Holinger
tation of the superior trachea. Other surgeons have reported good results using the slide tracheoplasty and cartilage tracheoplasty techniques. Intracardiac defects should be repaired simultaneously with the tracheal stenosis. The postoperative care of these patients is very complex and requires attention to detail and vigilance with regard to airway management. Emergency operative bronchoscopic management must be available 24 hours a day. In our series, patients who had dehiscence of a patch were managed successfully with vascularized muscle flaps brought into the mediastinum. We cannot emphasize enough the multidisciplinary approach to these patients and the close collaboration and coordination required between thoracic surgery, otolaryngology, anesthesiology, and pediatric intensive care unit services. The ongoing evaluation of our techniques for these infants will be the key to successful long-term outcomes.
References 1. Cantrell JR, Guild HG: Congenital stenosis of the trachea. Am J Surg 108:297-305, 1964 2. Janik JS, Nagaraj HS, Yacoub U, et al: Congenital funnelshaped tracheal stenosis: An asymptomatic lethal anomaly of early infancy. J Thorac Cardiovasc Surg 83:761-766, 1982 3. Greene DA: Congenital complete tracheal rings. Arch Otolaryngol 102:241-243, 1976 4. Benjamin B, Pitkin J, Cohen D: Congenital tracheal stenosis. Ann Otol Rhinol Laryngol 90:364-371, 1981 5. Kimura K, Mukohara N, Tsugawa C, et al: Tracheoplasty for congenital stenosis of the entire trachea. J Pediatr Surg 17:869-871, 1982 6. Idriss FS, DeLeon SY, Ilbawi MN, et al: Tracheoplasty with pericardial patch for extensive tracheal stenosis in infants and children. J Thorac Cardiovasc Surg 88:527536, 1984 7. Jonas RA, Spevak PJ, McGill T, et al: Pulmonary artery sling: Primary repair by tracheal resection. J Thorac Cardiovasc Surg 97:548-550, 1989 8. Tsang V, Murday A, Gillbe C, et al: Slide tracheoplasty for congenital funnel-shaped tracheal stenosis. Ann Thorac Surg 48:632-635, 1989 9. Elliott MJ, Haw MP, Jacobs JP, et al: Tracheal reconstruction in children using cadaveric homograft trachea. Eur J Cardiothorac Surg 10:707-712, 1996 10. Backer CL, Mavroudis C, Dunham ME, et al: Repair of congenital tracheal stenosis with a free tracheal autograft. J Thorac Cardiovasc Surg 115:869-874, 1998 11. Alboliras ET, Backer CL, Holinger LD, et al: Pulmonary
12.
13.
14.
15.
16.
17. 18.
19.
20.
21.
22.
23.
24.
25.
26.
artery sling–Diagnostic and management strategy. Pediatrics 98:530, 1996 (suppl) (abstr) Backer CL, Mavroudis C, Dunham ME, et al: Pulmonary artery sling: Results with median sternotomy, cardiopulmonary bypass, and reimplantation. Ann Thorac Surg 67:1738-1745, 1999 Backer CL, Pensler JM, Tobin GR, et al: Vascularized muscle flaps for life-threatening mediastinal wounds in children. Ann Thorac Surg 57:797-802, 1994 Furman RH, Backer CL, Dunham ME, et al: The use of balloon-expandable metallic stents in the treatment of pediatric tracheomalacia and bronchomalacia. Arch Otolaryngol Head Neck Surg 125:203-207, 1999 Backer CL, Mavroudis C, Dunham ME, et al: Reoperation after pericardial patch tracheoplasty. J Pediatr Surg 32: 1108-1111, 1997 Dayan SH, Dunham ME, Backer CL, et al: Slide tracheoplasty in the management of congenital tracheal stenosis. Ann Otol Rhinol Laryngol 106:914-919, 1997 Grillo HC: Slide tracheoplasty for long-segment congenital tracheal stenosis. Ann Thorac Surg 58:613-621, 1994 Dodge-Khatami A, Backer CL, Crawford SE, et al: Topical vascular endothelial growth factor (VEGF) enhances free tracheal autograft healing in an experimental rabbit model of tracheal reconstruction. Surg Forum 50:146-147, 1999 Heinemann MK, Ziemer G, Sieverding L, et al: Longsegment tracheal resection in infancy utilizing extracorporeal circulation, in Imai Y, Momma K (eds): Proceedings of the 2nd World Congress of Pediatric Cardiology and Cardiac Surgery. New York, Futura Publishing, 1998, pp 711-713 Cotter CS, Jones DT, Nuss RC, et al: Management of distal tracheal stenosis. Arch Otolaryngol Head Neck Surg 125:325-328, 1999 Tsugawa C, Kimura K, Muraji T, et al: Congenital stenosis involving a long segment of the trachea: Further experience in reconstructive surgery. J Pediatr Surg 23: 471-475, 1988 Jaquiss RDB, Lusk RP, Spray TL, et al: Repair of longsegment tracheal stenosis in infancy. J Thorac Cardiovasc Surg 110:1504-1512, 1995 Kamata S, Usui N, Ishikawa S, et al: Experience in tracheobronchial reconstruction with a costal cartilage graft for congenital tracheal stenosis. J Pediatr Surg 32:54-57, 1997 Chahine AA, Tam V, Ricketts RR: Use of the aortic homograft in the reconstruction of complex tracheobronchial tree injuries. J Pediatr Surg 34:891-894, 1999 Browdie DA, Bernstein RV, Johnson R: Materials for tracheoplasty: Which work? Which are best? J Thorac Cardiovasc Surg 113:810, 1997 (letter) Jacobs JP, Elliott MJ, Haw MP, et al: Pediatric tracheal homograft reconstruction: A novel approach to complex tracheal stenoses in children. J Thorac Cardiovasc Surg 112:1549-1558, 1996
T
he most frequent congenital tracheal anomaly requiring operative intervention is tracheal stenosis secondary to complete tracheal rings. This occurs when there is absence of the posterior membranous trachea and the cartilage of the trachea is circumferential (Fig 1). The tracheal orifice in these patients may be exceedingly small and cause presentation with severe respiratory distress within the first several weeks of life. Tracheal stenosis has been categorized by Cantrell and Guild1 into three major subsets; segmental (localized) stenosis, funnel-like stenosis, and generalized hypoplasia (long-segment stenosis). In the late 1970s and early 1980s, most case reports regarding this lesion reported the
From the Division of Cardiovascular-Thoracic Surgery, the Division of Pediatric Otolaryngology, and the Department of Communicative Disorders, Children’s Memorial Hospital, and the Department of Surgery, Northwestern University Medical School, Chicago, IL. Address reprint requests to Carl L. Backer, MD, Division of Cardiovascular-Thoracic Surgery-M/C#22, Children’s Memorial Hospital, 2300 Children’s Plaza, Chicago, IL 60614. Copyright © 2002 by W.B. Saunders Company 1092-9126/02/0501-0006$35.00/0 doi: 10.1053/pcsu.2002.29718
diagnosis but had a fatal outcome.2,3 Benjamin et al’s4 review of the medical management of these patients reported a 43% mortality rate. Many of these patients have associated anomalies such as pulmonary artery sling and severe intracardiac congenital anomalies that contribute to the morbidity and mortality of this patient population. The first successful procedure for tracheal stenosis secondary to complete tracheal rings was reported by Kimura et al5 in 1982. They used a rib cartilage graft to perform an anterior tracheoplasty. This was followed in 1984 by the series reported by Idriss et al6 who performed tracheoplasty with an autologous pericardial patch for extensive tracheal stenosis. Jonas et al7 reported tracheal repair by tracheal resection in 1989. The slide tracheoplasty was also reported in that year by Tsang et al.8 Elliott et al9 reported the use of cadaveric homograft material for tracheal reconstruction in 1996. We reported the use of a free tracheal autograft in 1998.10 At Children’s Memorial Hospital (Chicago, IL), the surgical advance of pericardial tracheoplasty devised by Dr Idriss and the strong support from our otolaryngology division established our
Pediatric Cardiac Surgery Annual of the Seminars in Thoracic and Cardiovascular Surgery, Vol 5, 2002: pp 173-186
173
174
Backer, Mavroudis, and Holinger
service as a center of referrals for tracheal stenosis. This has allowed us to develop a comprehensive program combining the respective skills of cardiothoracic surgery and otolaryngology. Over the past 19 years we have used six different techniques for infants with primary or recurrent tracheal stenosis. These techniques include pericardial tracheoplasty, tracheal resection, cartilage tracheoplasty, tracheal autograft, slide tracheoplasty, and aortic homograft patch tracheoplasty. Our techniques have evolved over time as we have gained experience with the advantages and disadvantages of the different techniques. This review will examine the short-term and long-term outcomes of our patients with congenital tracheal stenosis and review the world literature with regard to tracheal surgery in infants and children.
Patients and Methods
Figure 1. Illustration of a patient with complete tracheal rings affecting approximately two thirds of the tracheal length. The upper cutaway shows a normal tracheal ring with the anterior cartilage and relatively flat posterior membranous trachea. In the midportion, a complete tracheal ring is shown which is a circumferential cartilage ring with no membranous trachea. The lumen at the site of the complete tracheal ring can be quite small. (From Backer, CL, Mavroudis C, Gerber ME, et al: Tracheal surgery in children: An 18-year review of four techniques. Eur J. Cardiothorac Surg, vol 19, pp 777-784, 2001. Reprinted with permission.)
Between 1982 and 2001, 54 consecutive children with complete tracheal rings have undergone surgical repair at Children’s Memorial Hospital (see Table 1). The objective of this chapter is to review the short-term and long-term outcomes of the six different techniques used in these patients. The patients’ ages ranged from 7 days to 72 months with a mean age of 6 months. There were 33 boys and 21 girls. Nineteen of these patients had an associated pulmonary artery sling; thirteen had a significant associated intracardiac anomaly. All procedures were performed with the use of cardiopulmonary bypass. Simultaneous repair of pulmonary artery sling and cardiac abnormalities was undertaken. Symptoms at the time of presentation included respiratory distress, apnea, cyanosis, and stridor. Often these
Table 1. Techniques Used for Tracheoplasty at Children’s Memorial Hospital (Chicago, IL) Technique
Primary Secondary (No. of Pts) (No. of Pts)
Pericardial patch Tracheal autograft Tracheal resection Slide tracheoplasty Cartilage tracheoplasty Aortic homograft patch
28 15 9 2 0 0
2
4 1
Total
54
7
Repair of Congenital Tracheal Stenosis
patients presented as “difficult intubations” at the time of attempted intubation for either other surgical procedures or for respiratory distress. Many of these patients were referred with an endotracheal tube quite small for the patient’s age placed just through the vocal cords. Several of these patients had extremely high CO2 levels at the time of presentation. All patients had rigid bronchoscopy, either a day or two before the operative intervention or at the same time. All recent patients have had transthoracic echocardiography to rule out a pulmonary artery sling.11 Children who required only an isolated tracheal operation were placed on cardiopulmonary bypass with an aortic cannula and a single atrial cannula. These patients were cooled to 32°C and remained in normal sinus rhythm throughout the procedure. If the patient had an associated pulmonary artery sling, the repair of the sling was performed while on cardiopulmonary bypass. This was performed before the tracheal repair to prevent contamination of the vascular anastomosis by the open tracheal lumen. The technique of pulmonary artery sling repair was to transect the left pulmonary artery from its origin on the right pulmonary artery, dissect the left pulmonary artery free of the posterior mediastinal attachments behind the trachea and anterior to the esophagus, and move the left pulmonary artery anterior to the trachea. The left pulmonary artery was then reimplanted into the main pulmonary artery at a site that approximated where the normal left pulmonary artery would have originated.12 This was typically either adjacent to or at the site of a ligated ligamentum or patent ductus arteriosus. The left pulmonary artery anastomosis was commonly performed with interrupted PDS II (polydioxanone; Ethicon, Inc, Somerville, NJ) or Prolene (polypropylene; Ethicon, Inc) suture. If there was an associated intracardiac lesion that required repair, bicaval venous cannulation was used. The intracardiac lesion was repaired during a period of aortic cross clamp time before the tracheal repair, again, attempting to avoid contamination of any intracardiac material such as Dacron or Gore-Tex (W.L. Gore & Assoc, Flagstaff, AZ) by the open tracheal lumen.
Pericardial Patch Tracheoplasty This operation was our procedure of choice between 1982 and 1995. Immediately following the
175
median sternotomy, an autologous pericardial patch is harvested. This patch extends from the diaphragmatic surface to the reflection of the pericardium at the great vessels, and is typically 5 to 6 cm in length and 2 to 3 cm in width. The anterior surface of the trachea is dissected extensively from the cricoid to the carina. After cardiopulmonary bypass has been initiated, a number 11 blade is used to incise the trachea anteriorly throughout the extent of the complete tracheal rings. In the early part of our experience, we performed this under bronchoscopic guidance. Most recently we used direct visualization of the complete tracheal rings from anteriorly to guide the extent of this incision. The pericardial patch is brought onto the field and sutured in place using interrupted 6-0 Vicryl sutures (polyglactin 910; Ethicon, Inc) (Fig 2). By starting the suturing at the most inferior aspect of the incision adjacent to the carina, the accumulation of blood and secretions in the tracheal lumen is minimized. We used a running suture technique for only one patient in this series. That patient developed a patch dehiscence and lethal pseudomonas mediastinitis. Following placement of the patch, rigid bronchoscopy is performed as cardiopulmonary bypass is continued. This confirms relief of stenosis and is used to suction the tracheobronchial tree clear of blood and secretions. The endotracheal tube is then reintroduced so that the tip is just above the carina. The endotracheal tube effectively stents the patch open. Two small hemoclips are placed at the superior and inferior aspects of the patch. This helps to radiographically identify the extent to the patch and its location vis-a`-vis the endotracheal tube in the postoperative period. The mediastinum is filled with saline and the patient is ventilated to a peak airway pressure of 40 cm H2O to assess for any leaks in the patch. Leaks are sutured closed with interrupted Vicryl suture. The patient is then ventilated, warmed, and weaned from cardiopulmonary bypass. In the current era, we perform modified ultrafiltration for 10 minutes. The patch is sealed with Tisseel glue (Baxter Healthcare Corp, Glendale, CA). Standard chest closure is performed. The patient is kept paralyzed, intubated, and ventilated for 1 week. At that time, bronchoscopic assessment is performed to look for healing of the patch, sites of residual stenosis, and to remove granulation
176
Backer, Mavroudis, and Holinger
ing of the patch and to remove granulation tissue as required.
Tracheal Autograft
Figure 2. Illustration of a completed pericardial patch tracheoplasty. The pericardial patch has been sutured in the anterior tracheal opening. The pericardium is anchored with interrupted Vicryl sutures. In cross-section (inset), the tracheal lumen is shown now to be normal in size. (From Backer CL, Mavroudis C, Gerber ME, et al: Tracheal surgery in children: An 18-year review of four techniques. Eur J Cardiothorac Surg, vol 19, pp 777-784, 2001. Reprinted with permission.) tissue. Over the next 1 to 2 weeks the patient is weaned from the ventilator and is usually extubated 2 to 3 weeks after the procedure. Serial bronchoscopies are performed to assess the heal-
Tracheal autograft has been our procedure of choice since 1996. The tracheal autograft technique is performed with a median sternotomy approach and the use of cardiopulmonary bypass. The entire trachea is extensively mobilized circumferentially. This allows eventual resection of a significant portion of the midtrachea. The isthmus of the thyroid gland is divided. Dissection is carried below the carina well onto the right and left mainstem bronchi. Following dissection, the anterior trachea is incised in the area of the complete tracheal rings with a number 11 blade (Fig 3, left). With the tracheal lumen open through the rings, an assessment is made as to how much trachea can be resected without causing excessive tension at the eventual posterior anastomosis (see Figs 3, center and right). The tracheal autograft is harvested by excising the midportion of the open trachea. The autograft is usually 2 cm in length. The two remaining ends of the trachea are anastomosed posteriorly with multiple interrupted 6-0 PDS sutures (Fig 4, left). The sutures are placed to keep the sutures and knots out of the tracheal lumen. This helps to diminish the incidence of granulation tissue formation at the suture lines. The tracheal autograft that was previously harvested is used to patch the remaining anterior opening in the trachea. The corners of the autograft are trimmed (Fig 4, center). Sometimes the autograft has to be made smaller by cutting 1 to 2 mm along the length of each side of the autograft. The autograft is anchored in place with interrupted 6-0 PDS sutures (Fig 4, right). If the autograft is not long enough to augment the existing opening in the trachea, a composite with pericardium is made and the inferior portion of the anterior tracheal opening is patched with the autograft. The autograft is trimmed in a different fashion, not cutting the corners of the superior aspect of the autograft. The remaining upper opening in the trachea is then patched with fresh autologous pericardium, again anchored with 6-0 PDS sutures (Fig 5). This, of necessity, results in an anastomosis between pericardium and the tracheal autograft. Again, the suture line is tested for air leaks by ventilating the patient to 40 cm H2O pressure. Tisseel glue is again used to seal
Repair of Congenital Tracheal Stenosis
177
Figure 3. Tracheal autograft. An anterior longitudinal incision is performed through the complete extent of the tracheal rings (left). The midportion of the trachea is excised to be used as the tracheal autograft (center). The two remaining orifices of the trachea are reapproximated posteriorly (right). (Reprinted with permission.10)
the suture line. Two or three hemoclips are placed to identify the upper and lower aspect of the autograft or the upper and lower aspect sequentially of the pericardium and the autograft. The strategy for endotracheal tube placement depends on the type of autograft. If there is no pericardial augmentation, the endotracheal tube is kept above the area of the autograft. If the pericardium is used, the midportion of the tube is placed at the midpoint of the tracheal autograft. The tube will then stent the pericardium open. In our series, no patient had a neck brace or a chin stitch. The patient is kept paralyzed and ventilated for 3 to 5 days, at which time follow-up bronchoscopy is performed. If the tube is above the area of the anastomoses, this can be done with flexible bronchoscopy. If the tracheal lumen is patent with minimal granulation tissue and the child is stable from a respiratory and hemodynamic standpoint, they are weaned from the ven-
tilator over the next 3 to 5 days. Bronchoscopy is again performed just before extubation and then electively just before hospital discharge and at 3- and 6-months postdischarge. Some patients who have formation of granulation tissue or residual stenosis may require bronchoscopy more frequently.
Tracheal Resection Tracheal resection is applicable only to patients with short-segment (localized) tracheal stenosis. In our series, all tracheal resections have been performed with the use of cardiopulmonary bypass. The trachea is mobilized in a fashion similar to that described for the tracheal autograft. If there is any question about the site of the tracheal stenosis, rigid bronchoscopy is performed during cardiopulmonary bypass and a 25-gauge needle is used to penetrate the trachea. The needle is identified endoscopically and used to
178
Backer, Mavroudis, and Holinger
Figure 4. Tracheal autograft. The posterior anastomosis is performed with interrupted PDS sutures (left). The autograft is prepared by trimming the corners of the autograft (center). The autograft is sutured in place to cover the remaining opening in the anterior trachea (right). (Reprinted with permission.10)
guide the surgeon as to the length of the tracheal resection. In most cases, an anterior tracheotomy is made in a longitudinal fashion with a number 11 blade through the area of the stenotic rings. This confirms the location of the rings and allows one to convert to the tracheal autograft technique should the length of tracheal stenosis be longer than expected from the preoperative evaluation. In most cases a bilateral hilar release, freeing of the pulmonary artery from the pericardium, and in some instances a hyoid release are used to provide enough mobility to perform a successful end-to-end anastomosis. This anastomosis is performed with interrupted 6-0 PDS suture (Fig 6). In most cases, the endotracheal tube is positioned above the anastomosis; however, in some cases the tube was passed through the anastomosis. These patients were kept paralyzed for a much shorter period of time, 24 to 48 hours, and then were usually weaned from the ventilator and extubated. Most patients had one bronchoscopy just before extubation and one just before hospital discharge.
Slide Tracheoplasty We used slide tracheoplasty for two patients in 1996. Again, the patient is approached through a median sternotomy with the use of cardiopulmonary bypass. Slide tracheoplasty requires mobilization of the trachea in its entirety, similar to the autograft and tracheal resection techniques. The exact midportion of the stenosis must be identified because the trachea must be transected at this point (Fig 7). We have used bronchoscopic identification to guide us to the midpoint of the tracheal stenosis, despite what the external appearance of the trachea might be. The superior portion of the trachea is then incised posteriorly through the extent of the complete tracheal rings. In a similar fashion, the inferior aspect of the trachea is incised anteriorly, again through the extent of the complete tracheal rings. At both ends of the tracheal transection the corners are trimmed so that the leading edge will fit into the V-portion of the other component (Fig 8). The two components are slid together and anastomosed with interrupted 6-0 PDS sutures (Fig 9).
Repair of Congenital Tracheal Stenosis
179
Figure 5. Composite tracheal autograft. (A) The posterior anastomosis is performed. (B) The autograft is trimmed, only cutting the inferior corners of the autograft. A portion of pericardium is harvested and tailored. (C) The autograft is sutured in place anteriorly adjacent to the carina. The pericardial patch is inserted superiorly to complete the repair. (From Backer CL, Mavroudis C, Gerber ME, et al: Tracheal surgery in children: An 18-year review of four techniques. Eur J Cardiothorac Surg, vol 19, pp 777-784, 2001. Reprinted with permission.)
The completed anastomosis results in a trachea that is half as long and has four times the luminal diameter (Fig 10). The endotracheal tube is positioned in the midportion of the trachea. We and others8 have managed these patients postoperatively in a fashion very similar to those patients who have undergone a tracheal resection.
Cartilage Tracheoplasty Cartilage tracheoplasty has been described by several surgeons as a primary strategy for the patient with complete tracheal rings. In our series, however, we have used it only as a secondary operation after pericardial tracheoplasty. These patients typically have had distal tracheal stenosis and were patients who presented both with complete tracheal rings and a pulmonary artery sling. In these patients the pulmonary artery
sling affected the distal trachea, causing severe distal tracheal stenosis and often right mainstem bronchial stenosis. These patients were all operated on through a repeat sternotomy with the use of cardiopulmonary bypass. The cartilage graft was harvested from the fourth or fifth anterior cartilage on the right side. This cartilage is excised in toto and then cut longitudinally in half. Our otolaryngology colleagues, who commonly use these grafts for laryngotracheal reconstruction in the neck, then assist with the cartilage shaping. This results in a fairly thin cartilage with a step at the edge for the insertion of the cartilage into the tracheal lumen. The distal trachea is incised under bronchoscopic guidance through the area of stenosis. This was typically from the distal trachea out onto the right main-
180
Backer, Mavroudis, and Holinger
autograft became infected with pseudomonas, and during the time of an emergent bronchoscopy four days postoperatively, the autograft was disrupted. The child had emergency resternotomy, placement on cardiopulmonary bypass, and removal of the autograft. A cryopreserved aortic homograft was then used for the tracheal reconstruction. The aortic homograft was cut in a fashion similar to the patch we previously had used as a pericardial patch. After inserting the homograft in place with interrupted PDS sutures, the pectoralis major muscle was brought through an intercostal window to treat the child’s mediastinitis.13 The child had a very stormy postoperative course requiring balloon expandable Palmaz tracheal stents (Cordis Corp, Miami, FL)
Figure 6. Tracheal resection. The localized segment of tracheal stenosis is excised. The trachea is extensively mobilized, specifically mobilized in the right and left mainstem bronchus and carina. This allows the trachea to be brought together without tension for an end-toend anastomosis using interrupted PDS sutures.
stem bronchus. The cartilage is then fixed in place with interrupted Vicryl or PDS sutures (Fig 11). The patients were managed postoperatively with the endotracheal tube in the midportion of the trachea, similar to a patient after tracheal resection.
Aortic Homograft One patient in our series, following tracheal autograft, had dehiscence of the autograft patch. This was a 2-month-old boy who underwent complete repair of pulmonary artery sling, tetralogy of Fallot, and had a tracheal autograft. The child had a tracheal right upper lobe bronchus. This is the only child in the series in whom the autograft was divided in half longitudinally. These halves were used separately to patch the tracheal opening in an effort not to use pericardium. The
Figure 7. Slide tracheoplasty. The trachea is transected at the midpoint of the long-segment congenital tracheal stenosis. The superior portion of the trachea is incised posteriorly, the inferior aspect of the trachea is incised anteriorly. (Reprinted with permission.16)
Repair of Congenital Tracheal Stenosis
181
Figure 8. Slide tracheoplasty. The upper trachea has been opened posteriorly, the lower trachea has been opened anteriorly. The corners of the transected trachea are trimmed so that the leading edge will fit into the Vportion of the other component. (Reprinted with permission.16)
at 4 weeks and 3 months postoperatively. He had a tracheostomy at 7 months postoperatively. His hospital course was nearly 11 months and then he was discharged home. During bronchoscopy at the referring institution 6 months later the trachea was perforated and the patient died.
Results In this series of 61 operations on 54 children, there were three early deaths; two after pericardial tracheoplasty and one after tracheal autograft. There were eight late deaths, five after pericardial tracheoplasty, one after tracheal autograft, one after slide tracheoplasty, and one after tracheal resection. Nineteen patients had a pulmonary artery sling and 13 patients had a significant intracardiac anomaly. The associated intracardiac anomalies are shown in Table 2. The number of complete tracheal rings involved ranged from two to 18 rings (mean 14 ⫾ 5). Nine patients had a tracheal right upper lobe bronchus. Eight patients had a patent ductus arteriosus, three had an absent right lung, two an absent left lung, and two a severely hypoplastic right lung. In this series there was no complication specifically related to the use of cardiopulmonary bypass and no patient required a reop-
Figure 9. Slide tracheoplasty. This illustrates the beginning of the long anastomotic suture line that is performed with interrupted 6-0 PDS suture while the patient is on cardiopulmonary bypass. (Reprinted with permission.16)
182
Backer, Mavroudis, and Holinger
Table 2. Associated Significant Cardiac Anomalies Anomaly
Number
Tetralogy of Fallot Ventricular septal defect Complete atrioventricular septal defect Pulmonary atresia-Ventricular septal defect Double outlet right ventricle Atrial septal defect
4 4 2 1 1 1
Total
13
tracheoplasty. Follow-up is complete in all patients. Figure 10. Anterior (left) and lateral views (right) of the anastomosed trachea. The tracheal length has been reduced by almost half and the internal luminal diameter has been increased by four times. The cross-sectional appearance of the trachea is shown in the inset. (Reprinted with permission.16) eration for bleeding. The median length of hospital stay was 60 days for pericardial tracheoplasty, 28 days for tracheal autograft, 14 days for tracheal resection, and 18 days for the slide
Pericardial Tracheoplasty Mean age of the patients in this group was 6.5 ⫾ 5.5 months; median age was 6 months. This operation was performed between 1982 and 1995 as a primary procedure on 28 patients and as a secondary procedure on two patients. There were two early deaths. One child died 2 weeks postoperatively after developing pseudomonas mediastinitis from an air leak from the patch. This child was the only patient in this series where a running suture technique was used for the patch. The second death occurred in a 2-week-old infant
Figure 11. Cartilage tracheoplasty. The distal tracheobronchial stenosis has been opened by cutting through the pericardial patch and out onto the right mainstem bronchus. The cartilage craft is carved, inserted into the tracheal opening with the perichondrium facing the lumen, and anchored with interrupted PDS suture. (Reprinted with permission.15)
Repair of Congenital Tracheal Stenosis
with an absent right lung. This child had severe pulmonary hypertension following the procedure and required extracorporeal membrane oxygenation support in the operating room. The child died of an intraventricular hemorrhage 1 week after the operation. There were five late deaths after pericardial tracheoplasty. Two patients died of complications of perforation of the trachea at the time of bronchoscopy at the referring institution (7 and 12 months postoperatively). One child died the day after a tracheostomy 5 months postoperatively. One child died 15 months postoperatively of pulmonary hypertension; that child had also had atrioventricular septal defect repair at the time of tracheoplasty. The final late death was in a child with multiple anomalies who had colonic perforation 22 months following pericardial tracheoplasty. Six of the patients in the pericardial tracheoplasty series required reoperation for recurrent tracheal stenosis. Two patients had a second pericardial patch inserted, four patients had placement of a rib cartilage graft between 2 and 6 months after the pericardial patch. Three patients required five balloon expandable Palmaz stents.14 Six patients required a tracheostomy. In this subgroup, significant risk factors for reoperation included younger age at initial repair, associated pulmonary artery sling, and tracheal right upper lobe bronchus.15
Tracheal Autograft The tracheal autograft operation was first performed in January 1996 and has now been used in 15 consecutive patients. Mean age of these patients was 6 ⫾ 6 months; median age was 4.5 months. Ten of the 15 patients were distant referrals. The mean number of complete tracheal rings was 15 rings. The length of the autograft ranged from 1.3 to 2.5 cm (mean, 2 cm). This was typically six to eight complete tracheal rings. Five of the 15 patients required a pericardial augmentation of the superior aspect of the reconstruction. In those patients, the mean length of the pericardial patch was 2.5 cm. There was one early death (7% early mortality) and there has been one late death (7% late mortality). The one early death occurred 26 days postoperatively in a 6-month-old infant who had previously had two sternotomies for palliation of tetralogy of Fallot. The child had multiple anomalies and required extracorporeal membrane oxygenation support
183
postoperatively because of low cardiac output. He died of sepsis and anasarca. Autopsy showed multiple myocardial infarctions and necrotizing pneumonia, but a widely patent healed tracheal autograft. The late death occurred as described previously in the patient who required an aortic homograft rescue procedure for a failed tracheal autograft. One other patient had significant postoperative complications consisting of recurrent tracheal stenosis related to the pericardial patch portion of the repair. That patient required a Palmaz stent at 3 months postoperatively and a tracheostomy at 4 months postoperatively. That patient had a prolonged (5 months) hospital course but is now 3 years postoperatively and doing well. Three patients required temporary tracheostomies. The other patients in the series are currently asymptomatic from a respiratory standpoint.
Tracheal Resection Nine patients have undergone tracheal resection for shorter segments of tracheal stenosis between 1990 and 2000. The mean age of these patients was 14 months; median age was 4 months. The number of complete tracheal rings excised ranged from two to eight rings with a mean of five rings. There was one late death following tracheal resection. None of these patients required a tracheostomy. Mean hospital stay was 14 days. The one late death occurred 2.5 months after repair of pulmonary artery sling and resection of complete tracheal rings. This child had multiple anomalies including an imperforate anus (status post colostomy), intestinal malrotation, and biliary atresia. The child had end-stage liver failure and was on the waiting list for a liver transplant.
Slide Tracheoplasty Two slide tracheoplasties were performed at our institution, both in 1996.16 One child was 3 months old, had 18 complete tracheal rings, and had simultaneous repair of double outlet right ventricle. Her hospital course was smooth and she was discharged 18 days postoperatively. The other child had pulmonary atresia and ventricular septal defect with 12 complete tracheal rings. She underwent simultaneous modified BlalockTaussig shunt placement and slide tracheoplasty. She developed recurrent tracheal stenosis and granulation tissue. The stenosis was secondary to a “figure-of-eight” configuration of the tracheal
184
Backer, Mavroudis, and Holinger
cross-section. The tracheal rings appeared to spring back to their original cross-sectional area. That child required a Palmaz tracheal stent and tracheostomy. She was hospitalized for 5 months and was unable to be weaned from the ventilator. She died of sepsis from aortic valve endocarditis.
Cartilage Tracheoplasty Four patients underwent reoperation for tracheal stenosis after pericardial tracheoplasty with a cartilage graft.15 One of these patients had a tracheostomy before and after the revision. Two other patients required tracheostomy after the revision. The number of days to hospital discharge ranged from 20 to 105, with a mean of 64 days. The first patient is currently asymptomatic and has a healed patent airway. The second patient died when the trachea was perforated during bronchoscopy at the referring institution 1 year postoperatively. The third patient required two Palmaz stents and a tracheostomy but is currently asymptomatic. The fourth patient required two stents, one in the trachea and one in the left main bronchus, and is currently asymptomatic with a tracheostomy.
Discussion Infants with tracheal stenosis secondary to complete tracheal rings present a significant challenge to the surgeon. These children often present with life-threatening respiratory distress and are admitted to the hospital in extremis. Management of these patients requires close collaboration between the pediatric cardiac surgeon and the pediatric otolaryngologist. The experience of our institution has been an evolution in the surgical procedures for these patients as we have gained experience with the different techniques and their attendant complications. Initially, pericardial tracheoplasty was the operation of choice at our institution and was used successfully for almost 15 years. Pericardial tracheoplasty had a low operative mortality, but because of difficulties with granulation tissue, patch tracheomalacia, and the attendant prolonged hospital course, we devised another technique that has become our standard procedure for these patients. In 1996 we attempted two slide tracheoplasties with mixed results. The first patient had a smooth postoperative course and recovery. The
second patient, however, had significant complications from granulation tissue and recurrent stenosis. In addition, there are some technical aspects of slide tracheoplasty that are difficult to deal with. The trachea must be transected at the exact midpoint of the stenosis otherwise the resultant tracheal openings will not correspond in length. When performing the slide tracheoplasty the upper trachea has to be lifted anteriorly for an initial series of posterior sutures that are difficult to see and to tie. Air leaks along this posterior suture line are particularly difficult to deal with.17 In late 1995 we became aware of the efforts of several groups in Europe who used tracheal homografts for rescue procedures in patients who had failure of a previous operation for congenital tracheal stenosis. Elliott et al9 reported the use of tracheal homograft for these critically ill patients. At this same time, we operated on a patient with an extremely long trachea and decided to resect a number of the tracheal rings, performing an end-to-end anastomosis posteriorly before the pericardial tracheoplasty. The removed tracheal rings were sent to pathology. After reading the report by Elliott et al, we were intrigued by the fact that the piece of trachea that had been discarded looked exactly like the tracheal homograft that they were using. Because of this, for the next patient in our series ( January 1996) who underwent repair of tracheal stenosis, we used the midportion of the trachea as a distal autograft. That child had a very diminutive trachea, pulmonary artery sling, and a ventricular septal defect. The postoperative course was very good and the child was discharged from the hospital 20 days postoperatively. Because of this initial success we had great enthusiasm for this technique and have now used it in the next 14 patients. Initially, we had fears that the autograft might necrose and that in patients who required pericardial augmentation there might be dehiscence of the segment between the autograft and the pericardium. No patient in our series has had either of these complications. Only one patient has had autograft dehiscence and we feel this was caused by the technical issues of making the lumen too small combined with a pseudomonas infection. In no patient was there nonhealing of the junction between the autograft and the pericardium. The autograft usually appears pale but pink, even
Repair of Congenital Tracheal Stenosis
during the initial bronchoscopy. By 5 to 7 days postoperatively, it almost appears to have normal pink mucosal color. Several of the patients have had a very dusky appearing autograft for the first one or two bronchoscopies, but again, none have undergone autograft lysis. In our laboratory, Dr Dodge Khatami performed a variation of the autograft operation on 30 rabbits.18 We had no autograft dehiscence or ischemia in these rabbits. In that model we showed that the topical application of vascular endothelial growth factor enhanced the healing of the autograft. The use of vascular endothelial growth factor resulted in reduced luminal stenosis, reduced mucosal fibrosis, and reduced inflammatory infiltrate. The tracheal architecture was better preserved in the vascular endothelial growth factor-treated animals, and they had a greater microvascular density of the healing tracheal wall. Currently, the tracheal autograft is our procedure of choice for patients with long-segment tracheal stenosis. The advantages of the tracheal autograft include use of all autologous material for the repair, technical ease of performance with all-anterior suturing, already-present epithelial lining of the autograft, intrinsic maintenance of the cartilage contour, potential for growth, and ready availability because the trachea in these infants often appears excessively long. The technique of tracheal resection works extremely well with infants with shorter segment tracheal stenosis. Jonas et al7 reported success with this technique in two patients in 1989. Heinemann et al19 reported success in nine of 10 infants who had resection of tracheal stenosis using extracorporeal circulation. Cotter et al20 recently reported on six patients who all underwent resection with end-to-end anastomosis with excellent results. In our current practice we limit the use of tracheal resections to patients with eight or less rings of tracheal stenosis. Extensive mobilization of the trachea allows the tracheal anastomosis to be performed without tension. These patients have short time to extubation and a shorter hospital stay than many of the other techniques in our series. In the world literature the operative mortality for tracheal resection in infants is approximately 8%.7,19,20 The use of cartilage grafts has been reported by several surgeons to be a successful tech-
185
nique.5,21,22 However, other surgeons have reported a rather high mortality for these patients.23 Our personal opinion is that cartilage is not as easily molded to the trachea as is pericardium or autograft for long-segment stenosis. It tends to maintain its natural torque that does not necessarily conform to the longitudinal tracheal opening. Because of the rigidity of the cartilage it is sometimes difficult to obtain an airtight seal. We have had the experience where the cartilage seems to “fall” into the trachea and actually obstructs the tracheal lumen. The use of aortic homograft has been successful both for primary repair of long-segment congenital tracheal stenosis and for repair of the trachea after extensive injury from trauma.24,25 Aortic homograft is easily available in most pediatric cardiac surgical centers, thus use of the aortic homograft should be a part of any surgeon’s potential armamentarium for these patients. However, because of problems similar to those associated with the use of pericardium, ie, patch tracheomalacia, granulation tissue, and the use of foreign material, we tend to avoid aortic homograft except for emergency situations. Although Elliott and Jacobs have reported good results with tracheal homograft, they have continued to limit this to patients undergoing rescue operation after failed prior procedures.9,26 The tracheal homograft has been reported to result in erosion into surrounding structures and probably should still be considered only as a rescue operation. The infant with tracheal stenosis secondary to complete tracheal rings is a challenge to the managing surgeon. These infants require close cooperation between pediatric cardiac surgeons and pediatric otolaryngologists. Diagnosis of these patients is achieved by rigid bronchoscopy. We have used echocardiography to rule out associated pulmonary artery sling (present in nearly one third of these patients). We recommend repair of the tracheal stenosis through a median sternotomy approach with the use of cardiopulmonary bypass. We currently recommend tracheal resection and end-to-end anastomosis for those patients who have eight complete tracheal rings or less. For patients with long-segment congenital tracheal stenosis, we recommend repair with a tracheal autograft technique. This can be performed with or without pericardial augmen-
186
Backer, Mavroudis, and Holinger
tation of the superior trachea. Other surgeons have reported good results using the slide tracheoplasty and cartilage tracheoplasty techniques. Intracardiac defects should be repaired simultaneously with the tracheal stenosis. The postoperative care of these patients is very complex and requires attention to detail and vigilance with regard to airway management. Emergency operative bronchoscopic management must be available 24 hours a day. In our series, patients who had dehiscence of a patch were managed successfully with vascularized muscle flaps brought into the mediastinum. We cannot emphasize enough the multidisciplinary approach to these patients and the close collaboration and coordination required between thoracic surgery, otolaryngology, anesthesiology, and pediatric intensive care unit services. The ongoing evaluation of our techniques for these infants will be the key to successful long-term outcomes.
References 1. Cantrell JR, Guild HG: Congenital stenosis of the trachea. Am J Surg 108:297-305, 1964 2. Janik JS, Nagaraj HS, Yacoub U, et al: Congenital funnelshaped tracheal stenosis: An asymptomatic lethal anomaly of early infancy. J Thorac Cardiovasc Surg 83:761-766, 1982 3. Greene DA: Congenital complete tracheal rings. Arch Otolaryngol 102:241-243, 1976 4. Benjamin B, Pitkin J, Cohen D: Congenital tracheal stenosis. Ann Otol Rhinol Laryngol 90:364-371, 1981 5. Kimura K, Mukohara N, Tsugawa C, et al: Tracheoplasty for congenital stenosis of the entire trachea. J Pediatr Surg 17:869-871, 1982 6. Idriss FS, DeLeon SY, Ilbawi MN, et al: Tracheoplasty with pericardial patch for extensive tracheal stenosis in infants and children. J Thorac Cardiovasc Surg 88:527536, 1984 7. Jonas RA, Spevak PJ, McGill T, et al: Pulmonary artery sling: Primary repair by tracheal resection. J Thorac Cardiovasc Surg 97:548-550, 1989 8. Tsang V, Murday A, Gillbe C, et al: Slide tracheoplasty for congenital funnel-shaped tracheal stenosis. Ann Thorac Surg 48:632-635, 1989 9. Elliott MJ, Haw MP, Jacobs JP, et al: Tracheal reconstruction in children using cadaveric homograft trachea. Eur J Cardiothorac Surg 10:707-712, 1996 10. Backer CL, Mavroudis C, Dunham ME, et al: Repair of congenital tracheal stenosis with a free tracheal autograft. J Thorac Cardiovasc Surg 115:869-874, 1998 11. Alboliras ET, Backer CL, Holinger LD, et al: Pulmonary
12.
13.
14.
15.
16.
17. 18.
19.
20.
21.
22.
23.
24.
25.
26.
artery sling–Diagnostic and management strategy. Pediatrics 98:530, 1996 (suppl) (abstr) Backer CL, Mavroudis C, Dunham ME, et al: Pulmonary artery sling: Results with median sternotomy, cardiopulmonary bypass, and reimplantation. Ann Thorac Surg 67:1738-1745, 1999 Backer CL, Pensler JM, Tobin GR, et al: Vascularized muscle flaps for life-threatening mediastinal wounds in children. Ann Thorac Surg 57:797-802, 1994 Furman RH, Backer CL, Dunham ME, et al: The use of balloon-expandable metallic stents in the treatment of pediatric tracheomalacia and bronchomalacia. Arch Otolaryngol Head Neck Surg 125:203-207, 1999 Backer CL, Mavroudis C, Dunham ME, et al: Reoperation after pericardial patch tracheoplasty. J Pediatr Surg 32: 1108-1111, 1997 Dayan SH, Dunham ME, Backer CL, et al: Slide tracheoplasty in the management of congenital tracheal stenosis. Ann Otol Rhinol Laryngol 106:914-919, 1997 Grillo HC: Slide tracheoplasty for long-segment congenital tracheal stenosis. Ann Thorac Surg 58:613-621, 1994 Dodge-Khatami A, Backer CL, Crawford SE, et al: Topical vascular endothelial growth factor (VEGF) enhances free tracheal autograft healing in an experimental rabbit model of tracheal reconstruction. Surg Forum 50:146-147, 1999 Heinemann MK, Ziemer G, Sieverding L, et al: Longsegment tracheal resection in infancy utilizing extracorporeal circulation, in Imai Y, Momma K (eds): Proceedings of the 2nd World Congress of Pediatric Cardiology and Cardiac Surgery. New York, Futura Publishing, 1998, pp 711-713 Cotter CS, Jones DT, Nuss RC, et al: Management of distal tracheal stenosis. Arch Otolaryngol Head Neck Surg 125:325-328, 1999 Tsugawa C, Kimura K, Muraji T, et al: Congenital stenosis involving a long segment of the trachea: Further experience in reconstructive surgery. J Pediatr Surg 23: 471-475, 1988 Jaquiss RDB, Lusk RP, Spray TL, et al: Repair of longsegment tracheal stenosis in infancy. J Thorac Cardiovasc Surg 110:1504-1512, 1995 Kamata S, Usui N, Ishikawa S, et al: Experience in tracheobronchial reconstruction with a costal cartilage graft for congenital tracheal stenosis. J Pediatr Surg 32:54-57, 1997 Chahine AA, Tam V, Ricketts RR: Use of the aortic homograft in the reconstruction of complex tracheobronchial tree injuries. J Pediatr Surg 34:891-894, 1999 Browdie DA, Bernstein RV, Johnson R: Materials for tracheoplasty: Which work? Which are best? J Thorac Cardiovasc Surg 113:810, 1997 (letter) Jacobs JP, Elliott MJ, Haw MP, et al: Pediatric tracheal homograft reconstruction: A novel approach to complex tracheal stenoses in children. J Thorac Cardiovasc Surg 112:1549-1558, 1996