- Email: [email protected]
Intermediate-term results of the free tracheal autograft for long segment congenital tracheal stenosis
Journal of Pediatric Surgery VOL 35, NO 6
JUNE 2000
Intermediate-Term Results of the Free Tracheal Autograft for Long Segment Congenital Tracheal Stenosis By Carl L. Backer, Constantine Mavroudis, Michael E. Dunham, and Lauren Holinger Chicago, Illinois
Background/Purpose: Since 1996 a new procedure—the free tracheal autograft—has been used to repair long segment congenital tracheal stenosis (LSCTS). The purpose of this report is to examine the intermediate-term results of that technique. Methods: Between January 1996 and July 1999, 10 infants underwent repair of LSCTS using a free tracheal autograft. Age ranged from 10 days to 23 months (mean age, 6.6 months). Six infants had a pulmonary artery (PA) sling; 5 had intracardiac anomalies. On cardiopulmonary bypass (CPB) the trachea was incised anteriorly through the area of stenosis. The midportion of the stenotic trachea was excised, and an end-to-end anastomosis was made posteriorly. The excised tracheal segment was trimmed and sutured in place anteriorly as a free autograft. In 5 patients the autograft was not long enough, and the upper portion of the tracheal opening was patched with pericardium.
required extracorporeal membrane oxygenation postoperatively for cardiac failure. The other 9 children are alive and well at 2 to 44 months postoperatively (mean follow-up, 24 months). One child had autograft dehiscence and required replacement of the autograft with an aortic homograft. Two children have tracheostomies at 6 and 36 months postoperatively. All children have had serial postoperative bronchoscopic examinations. Most recent bronchoscopies have shown widely patent tracheal lumina in all survivors.
Conclusion: Intermediate-term follow-up of children with a free tracheal autograft continues to support our use of this technique as our procedure of choice for infants with LSCTS. J Pediatr Surg 35:813-819. Copyright 娀 2000 by W.B. Saunders Company.
Results: There was 1 death 26 days postoperatively in a child that had simultaneous repair of tetralogy of Fallot and
INDEX WORDS: Tracheal stenosis, complete tracheal rings, tracheal autograft, tracheoplasty.
T
children that have undergone tracheal autograft for long segment congenital tracheal stenosis.
HE FIRST SUCCESSFUL repair of long segment congenital tracheal stenosis (LSCTS) at Children’s Memorial Hospital was with an anterior pericardial patch placed through a median sternotomy with the use of cardiopulmonary bypass in 1982.1 Over the next 14 years this procedure was used in 28 infants with a low operative mortality rate (7%) but a significant late mortality rate from airway complications (11%) and a relatively high reoperation rate (21%) for residual tracheal stenosis.2 In addition, these patients were noted to have frequent problems with granulation tissue requiring multiple bronchoscopies and tracheomalacia sometimes requiring prolonged intubation or stent placement.3,4 The tracheal autograft technique was used first essentially in desperation in an infant with an extremely dimunitive distal trachea, a pulmonary artery sling, and large ventricular septal defect (VSD).5 This report reviews our intermediate experience with infants and Journal of Pediatric Surgery, Vol 35, No 6 (June), 2000: pp 813-819
MATERIALS AND METHODS Between January 1996 and July 1999, 10 infants underwent repair of LSCTS using a free tracheal autograft technique. There were 7 boys and 3 girls. Age at surgery ranged from 10 days to 23 months (mean age, 6.6 months). Weight at surgery ranged from 2.9 to 10 kg (mean weight, 5.7 kg). All patients had cardiac echocardiography (ECHO), and most had a
From the Divisions of Cardiovascular-Thoracic Surgery and Otolaryngology, Northwestern University Medical School, and the Department of Surgery, The Children’s Memorial Hospital, Chicago, IL. Presented at the 1999 Annual Meeting of the American Academy of Pediatrics, Washington DC, October 8-10, 1999. Address reprint requests to Carl L. Backer, MD, Children’s Memorial Hospital, 2300 Children’s Plaza, m/c 22, Chicago, IL 60614. Copyright 娀 2000 by W.B. Saunders Company 0022-3468/00/3506-0001$03.00/0 doi:10.1053/js.2000.6847 813
814
BACKER ET AL
computed tomography (CT) scan with contrast. Six patients had a pulmonary artery (PA) sling. Five had intracardiac anomalies including tetralogy of Fallot (TOF, n ⫽ 2), ventricular septal defect (VSD), complete atrioventricular canal defect (CAVC), and atrial septal defect (ASD, n ⫽ 3). The number of complete tracheal rings ranged from 6 to 22 (mean, 15 rings). Five patients were transferred intubated and ventilated and required an urgent or emergent procedure. Seven of 10 patients were out-of-state referrals. Other associated congenital anomalies included patent ductus arteriosus (PDA, n ⫽ 5), tracheal right upper lobe (RUL) bronchus (n ⫽ 3), left superior vena cava (SVC, n ⫽ 2), subglottic stenosis (n ⫽ 2), hypoplastic right lung (n ⫽ 2), Down’s syndrome (n ⫽ 2), and absent left lung (n ⫽ 1). These patient characteristics are summarized in Table 1. All patients underwent rigid bronchoscopy within 2 days before the autograft or immediately after anesthetic induction for the autograft procedure. In most cases the bronchoscope itself could not be passed through the tracheal rings. In 3 cases not even the fine telescope (outer diameter, 2.5 mm) could be passed completely through the stenosis. Surgical approach was with a median sternotomy extended vertically onto the neck. Cardiopulmonary bypass was used in all cases to provide safe respiratory support while the trachea is being repaired. Bicaval cannulation (28°C) was used for cases in which an intracardiac anomaly was to be repaired, and right atrial cannulation (32°C) for all others. Intracardiac procedures were performed first, followed by PA sling repair, followed by tracheal repair. Intracardiac procedures in 5 patients included complete repair of VSD, ASD (n ⫽ 3), TOF (n ⫽ 2), and CAVC. In 6 patients, PA sling repair was performed. In 5 patients the left pulmonary artery (LPA) was transsected, moved anterior to the trachea, and reimplanted into the main pulmonary artery (MPA).6 In one patient with a hypoplastic right lung and a tiny right PA the left PA was translocated anterior to the trachea after tracheal division as originally described by Jonas et al.7 After intracardiac or PA sling repair, the entire trachea was extensively mobilized. This extended superiorly to the cricoid, and in most cases the isthmus of the thyroid gland was divided. Inferiorly, the dissection was carried below the carina and well onto the right and left mainstem bronchi. Bilateral hilar release and freeing of the PA from the pericardium also was included.8 The anterior trachea was incised with an 11 blade through the extent of the complete tracheal rings (Fig 1). An assessment was made as to how much trachea could be resected without causing excessive tension at the anastomosis. This excised midportion
Fig 1. Trachea with complete tracheal rings from ring 6 to just above the carina. Dotted line shows anterior incision to be made in trachea through the area of stenosis.
of the trachea is used for the free tracheal autograft (Fig 2). The 2 remaining ends of trachea are anastomosed posteriorly with multiple interrupted 6-0 PDS sutures (Ethicon Inc, Somerville, NJ) (Fig 3). These sutures are placed to keep the sutures and knots out of the tracheal lumen. The free tracheal autograft is used to patch the trachea anteriorly (Fig 3). This is accomplished by trimming the corners of the autograft and anchoring it in place with interrupted 6-0 PDS sutures (Fig 3). When the autograft can augment only the inferior portion of the anterior tracheal opening, the upper aspect of the tracheal incision is patched
Table 1. Patient Clinical Characteristics Patient No.
Age (mo)
Weight (kg)
1
2
4.5
2 3
5 2
4.9 4.2
4
0.3
2.9
5
12
10
6
8
6.2
7
6
5.0
8
6
6.0
9 23 9.9 10 2 3.5 Mean 6.6 ⫾ 6.7 5.7 ⫾ 2.4
Associated Anomalies
PA sling, VSD, LSVC, ASD, tracheal RUL Subglottic stenosis PA sling, PDA, hypoplastic right lung PA sling, PDA, LSVC, subglottic stenosis PA sling, hypoplastic right lung CAVC, PDA, Down’s syndrome TOF S/P shunt, S/P outflow patch, tracheal RUL PDA, Down’s syndrome, absent left lung PA sling, ASD PA sling, TOF, ASD, PDA
No. of Tracheal Rings
15 6 16 18 20 8 16 20 22 16 15.7 ⫾ 5.1
Fig 2. After the trachea has been opened anteriorly the midportion of the complete tracheal rings is excised to be used as the free tracheal autograft.
TRACHEAL AUTOGRAFT: INTERMEDIATE RESULTS
815
is stable, the paralysis is stopped, and the child is weaned from the ventilator over 3 to 5 days and extubated. Follow-up bronchoscopy is performed just before extubation, just before discharge, at 3 and 6 months, and then yearly.
RESULTS
Fig 3. The 2 ends of the trachea are anastomosed together posteriorly with interrupted sutures kept out of the lumen. The corners of the autograft are trimmed and the autograft is placed as an anterior patch with interrupted sutures.
with fresh autologous pericardium anchored with 6-0 PDS sutures (Fig 4). Pericardial augmentation was required in 5 patients. The mediastinum is copiously irrigated with an amikacin solution. Tissell glue (Baxter Healthcare Corp, Glendale, CA) is used to seal the suture lines on the trachea. Small hemoclips are placed in the soft tissue adjacent to the upper and lower aspects of the autograft for radiographic analysis of the position of the endotracheal (ET) tube. We try to keep the ET tube either (1) above the autograft if there is no pericardial augmentation or (2) in the midportion of the autograft in patients that have had pericardial augmentation. In patients in whom pericardial augmentation was used, the ET tube acts as a stent for the pericardial patch. No patient had a neck brace or chin stitch. The patient is kept paralyzed and ventilated for 3 to 5 days, at which time surveillance rigid bronchoscopy is performed. If the tracheal lumen is patent and the child
There was 1 early death (10%) and no late deaths. There was one significant complication—autograft dehiscence requiring replacement of the autograft with aortic homograft. Days of intubation ranged from 7 to 35 days (mean, 15 ⫾ 9.3 days), days of hospitalization ranged from 14 to 50 days (mean, 29.2 ⫾ 13.9 days). Most patients had 4 to 6 bronchoscopies during their hospitalization, although several ‘‘outliers’’ required multiple bronchoscopies. Currently, we are performing yearly bronchoscopies on all patients. Two patients still require tracheostomies at 6 and 36 months postrepair. Follow-up is complete in all patients and ranges from 2 to 44 months (mean, 24.1 ⫾ 16.4 months). Surgical findings, operative techniques, and clinical outcome are summarized in Table 2. The 1 early death occurred 26 days postoperatively in patient 7 in the series. This 6-month-old child had multiple congenital anomalies including tetralogy of Fallot, anomalous left coronary artery, agenesis of the left kidney, and extra digits. He had had 2 prior sternotomies for palliation of TOF with right ventricular outflow tract patch and modified Blalock-Taussig shunt. He underwent shunt takedown, complete repair of TOF, and autograft repair of 16 complete tracheal rings. He had progressive cardiac failure postoperatively and was placed on extracoporeal membrane oxygenation (ECMO) 12 hours after complete repair. He was supported on ECMO for 6 days. He became septic, and anasarca developed over a 2-week period. He then had multisystem organ failure and died 26 days postoperatively. Autopsy findings showed mulTable 2. Surgical Technique and Clinical Outcome Autograft Patient Length (cm) No.
1 2 3 4 5 6 7
Fig 4. For the infant with stenosis from just below the cricoid to the carina, the autograft is again taken from the midportion of the trachea. The autograft is placed inferiorly at the critical area for healing just above the carina. Note the autograft is trimmed preserving the upper corners. Pericardium anchored with interrupted sutures is placed superiorly to complete the tracheal reconstruction.
1.5 1.3 1.3 2.2 2.0 1.8 1.9
Pericardial Patch Length (cm)
Intubation (d)
Discharge (d)
Follow-Up (mo)
2.5 — 3.0 1.5 3.0 — —
13 20 44 7 14 38 7 19* 36 21 34 34 15 28 29 14 21 26 Died postoperative — — day 26 8 2.5 2.0 28 (Trach) 50 7 9 2.5 — 8 18 4 10† 2.0 — 35 49 2 Mean 1.9 ⫾ 0.4 2.4 ⫾ 0.6 15 ⫾ 9.3 29.2 ⫾ 13.9 24.1 ⫾ 16.4 *Required multiple readmissions over a 5-month period, eventually had a tracheostomy. †Had autograft dehisence and required emergent replacement of the autograft with an aortic homograft and later Palmaz stent.
816
BACKER ET AL
tiple areas of myocardial infarction and necrotizing bilateral pneumonia. The autograft repair was intact and without stenosis or evidence of infection. The child with autograft dehiscence underwent complete repair of PA sling, TOF, ASD, and autograft at 2 months of age. This child had a tracheal RUL bronchus and a stenotic ‘‘bridging bronchus.’’9 The autograft was divided in half longitudinally, and the 2 segments were both used for the anterior patch in an effort to not use pericardium. He had copious thick secretions that cultured positive for pseudomonas. On day 4 postoperation, during bronchoscopy, the autograft disrupted, and he had a massive air leak followed by desaturation and cardiac arrest. His chest was opened and the distal trachea was intubated directly. The upper autograft had lysed, and only the lower autograft could be identified. The patient was placed back on cardiopulmonary bypass (CPB) and the remaining autograft was removed. A cryopreserved aortic homograft was thawed and trimmed and placed as an anterior tracheal patch as described by Chahine et al.10 A vascularized muscle flap—the right pectoralis—was brought through an intercostal window to treat the pseudomonas mediastinitis.11 Because of collapse of the aortic homograft patch, a balloon expandable metallic (Palmaz) stent (20 mm length, 5 mm diameter) was placed 27 days after the reoperation.12 He was extubated 1 week later and discharged from the hospital 2 weeks after extubation. All patients currently are asymptomatic from a respiratory standpoint except for the 2 with tracheostomies. All have patent airways that on serial bronchoscopic examination appear to be growing normally with the patient. The autograft is difficult to differentiate from the surrounding trachea. The posterior end-to-end anastomosis has healed and has not been a source of restenosis. In 3 patients the pericardial portion of the repair has healed and presumably been replaced by ciliated pseudostratified columnar epithelium.13 In 2 other patients the pericardium has been the source of ‘‘patch tracheomalacia’’; and those 2 children have tracheostomies for that reason. DISCUSSION
We have reviewed our intermediate-term results with the technique of free tracheal autograft in 10 infants with long-segment congenital tracheal stenosis. The technique utilizes a partial tracheal resection, posterior end-to-end anastomosis, and anterior patching of the trachea with the resected piece—the tracheal autograft.5 Six of these patients had simultaneous repair of PA sling, and 5 had simultaneous repair of an intracardiac anomaly. Cardiopulmonary bypass was used in all patients. This greatly facilitated the tracheal diagnosis and repair. The telescope can be used in the trachea without ventilating the patient.
There is no risk of intraoperative asphyxia even with the tiniest tracheas (1.5 to 2.0 mm). The patency rate of the left pulmonary artery is greatly improved in pulmonary artery sling patients by using cardiopulmonary bypass.6 There were no complications secondary to cardiopulmonary bypass in this series. At a mean follow up of 24 months, 7 of 9 patients (78%) are essentially asymptomatic from a respiratory standpoint, and 2 are doing well with tracheostomies. We also have studied this technique in a laboratory model and have performed the tracheal autograft successfully in over 30 rabbits.14 The advantages of the autograft technique are several. The trachea in infants with LSCTS often is excessively long; hence, autograft ‘‘material’’ is readily available. There have been no complications related to the end-toend anastomosis and posterior suture line in this series. The autograft is already lined with respiratory epithelium. The cartilage intrinsically maintains the tracheal contour. The autograft has a potential for growth. In our laboratory model the autograft appears microscopically viable as early as 2 weeks after the repair.14 There seems to be minimal granulation tissue associated with the autograft especially at the critical junction of the patch with the carina. Our intermediate follow-up seems to indicate there is normal growth of the autograft. Other surgical alternatives for LSCTS include pericardial tracheoplasty,1,3,15 cartilage tracheoplasty,16-18 slide tracheoplasty,4,19,20-22 pulmonary homograft,23 and tracheal homograft.24 A summary of the number of patients treated with these techniques and the reported mortality rates are shown in Table 3. At our institution we used the pericardial tracheoplasty technique almost exclusively until 1996 (Fig 5). Although our operative mortality rate was low (7%), there was a significant late mortality from airway complications (11%) and a relatively high reoperation rate (21%) for residual stenosis.2 These patients required prolonged stenting of the trachea with the endotracheal tube and often required multiple bronchoscopies for dilation and removal of granulation tissue.3 Bando et al15 reported 12 patients having pericardial tracheoplasty with a mortality rate of 17%. Although we have changed to the tracheal autograft technique, we still use pericardium for the very long stenoses when the autograft is not long enough to patch the entire tracheal opening. By using a relatively short segment of pericardium in the upper trachea (away from the carina), the complications related to the use of pericardium are considerably fewer. The pericardium becomes replaced by ciliated pseudostratified columnar epithelium.13,15 We have used cartilage tracheoplasty successfully in 3 of 4 patients undergoing revision of pericardial tracheoplasty.2 Jaquiss et al17 reported excellent results with cartilage tracheoplasty, 6 patients (mean age, 3 months) with no mortality. Kamata et al18 reported quite different
TRACHEAL AUTOGRAFT: INTERMEDIATE RESULTS
817
Table 3. Published Results—Tracheoplasty Series Study
Technique
Year
Mean Age (mo)
No. of Patients
Mortality Rate (%)
Bando et al15 Backer et al2 Tsugawa et al16 Jaquiss et al17 Kamata et al18 Tsang et al19 Grillo8 Dayan et al4 Cunningham et al21 Lang et al22 Browdie et al23 Jacobs et al24 Backer et al (present study)
Pericardial tracheoplasty Pericardial tracheoplasty Cartilage tracheoplasty Cartilage tracheoplasty Cartilage tracheoplasty Slide tracheoplasty Slide tracheoplasty Slide tracheoplasty Slide tracheoplasty Slide tracheoplasty PA homograft Tracheal homograft Autograft
1996 1997 1988 1995 1997 1989 1994 1997 1998 1997 1997 1996 1999
7 6 10 3 5 6 10.5 yr 2 14 21 4 36 7
12 28 4 6 11 2 4 2 3 2 2 10 10
2 (17) 5 (18) 1 (25) 0 5 (45) 1 (50) 0 1 (50) 0 0 0 3 (30)* 1 (10)
*Nine of 10 had prior tracheal operation.
results, 11 patients (mean age, 5 months) with a 45% mortality rate. We have found cartilage (from rib) somewhat difficult to carve and implant. The cartilage is quite rigid and more difficult to achieve an air-tight seal. We and others have had the experience in which the cartilage is ‘‘too thick’’ and seems to ‘‘fall into the lumen,’’ occluding the lumen. Stokes et al25 reported on a patient in whom initial cartilage tracheoplasty from a rib had to be redone with ear cartilage for this very reason. The slide tracheoplasty technique was first reported in 2 patients by Tsang et al in 1989.19 In the past 10 years, 9 more patients have been reported with an overall mortality of 18%.8,4,21,22 Our experience comparing the 2 techniques has been that the autograft technique is easier to perform technically, is applicable to all patients with LSCTS (even those with complete tracheal rings from
Fig 5. Pericardial tracheoplasty. The trachea is incised anteriorly through the extent of the complete tracheal rings and patched anteriorly with autologous pericardium.
cricoid to carina), and results in an architecturally more desirable tracheal lumen. The slide tracheoplasty requires precise identification of the midportion of the stenosis before the initial tracheal transsection. This is not always easy to do either bronchoscopically or from external examination. If the trachea is transected initially in the wrong site, the 2 portions of trachea will not correspond and there will be a residual stenosis. 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 tie. In contrast, with the autograft technique, all the suturing is anterior and analogous to what surgeons are used to with the pericardial patch and cartilage tracheoplasty. Air leaks along the posterior suture line are particularly difficult to deal with. In the report of Grillo20 he noted having to perform an anterior vertical tracheal incision after the repair to deal with a persistent posterior air leak in an ‘‘inaccessible spot.’’ When the slide tracheoplasty is completed, the spring of the tracheal rings tends to form a ‘‘figure of 8’’ contour in cross section. This ‘‘unnatural’’ architectural configuration is a set-up for forces leading to granulation tissue and tracheomalacia. One of our patients had significant granulation tissue form at this waist and had flattening of the trachea requiring an endobronchial (Palmaz) stent and later a tracheostomy.4 The use of pulmonary homograft as an anterior tracheoplasty for infants with LSCTS was reported first by Browdie et al in 1997.23 Chahine et al10 reported the use of aortic homograft in 3 children, 1 with a gunshot wound, 1 with bronchial dehiscence after lung transplant, and 1 with an acquired tracheoesophageal fistula. We used this technique as rescue therapy for a child with autograft dehiscence. Although the homograft sealed well and with a muscle flap survived mediastinitis, it remained ‘‘floppy,’’ and patch tracheomalacia required an endobronchial (Palmaz) stent. Finally, Jacobs et al24 recently have reported the use of
818
BACKER ET AL
tracheal homograft for patients with tracheal stenosis from trauma, intubation, malignancy, and LSCTS. Their report details the use of a tracheal homograft in 10 infants with LSCTS, 9 of whom had a prior tracheal repair of some sort. Their survival rate of 70% in this complex group is extraordinary. We considered this option in the child in our series that had autograft dehiscence, but did not have a tracheal homograft available. Long segment congenital tracheal stenosis from com-
plete tracheal rings in children is often a life-threatening problem that frequently is associated with PA sling or major intracardiac anomalies. We recommend repair of the tracheal stenosis with the tracheal autograft technique and simultaneous repair of associated PA sling and cardiac anomalies. Care of these children after surgery is very demanding and requires close cooperation between pediatric otolaryngology and thoracic surgery departments.
REFERENCES 1. Idriss FS, DeLeon SY, Ilbawi MN, et al: Tracheoplasty with pericardial patch for extensive tracheal stenosis in infants and children. J Thorac Cardiovasc Surg 88:527-535, 1984 2. Backer CL, Mavroudis C, Dunham ME, et al: Reoperation after pericardial patch tracheoplasty. J Pediatr Surg 32:1108-1112, 1997 3. Dunham ME, Holinger LD, Backer CL, et al: Management of severe congenital tracheal stenosis. Ann Otol Rhinol Laryngol 103:351356, 1994 4. Dayan SH, Dunham ME, Backer CL, et al: Slide tracheoplasty in the management of congenital tracheal stenosis. Ann Otol Rhin Laryngol 106:914-919, 1997 5. 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 6. 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 7. Jonas RA, Spevak PJ, McGill T, et al: Pulmonary artery sling: primary repair by tracheal resection in infancy. J Thorac Cardiovasc Surg 97:548-550, 1989 8. Grillo HC: Surgical approaches to the trachea. Surg Gynecol Obstet 129:347-352, 1969 9. Gonzalez-Crussi FL, Padilla M, Miller JK, et al: Bridging bronchus: A previously undescribed airway anomaly. Am J Dis Child 130:1015, 1976 10. 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 11. Backer CL, Pensler J, Mavroudis C, et al: Vascularized muscle flaps for life threatening mediastinal wounds in children. Ann Thorac Surg 57:797-802, 1994 12. 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 13. Cheng ATL, Backer CL, Holinger LD, et al: Histopathological changes following pericardial patch tracheoplasty. Arch Otolaryngol Head Neck Surg 123:1069-1072, 1997
14. 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 vol L: 146-147, 1999 15. Bando K, Turrentine MW, Sun K, et al: Anterior pericardial tracheoplasty for congenital tracheal stenosis: Intermediate to long-term outcomes. Ann Thorac Surg 62:981-989, 1996 16. 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 17. Jaquiss RDB, Lusk RP, Spray TL, et al: Repair of long-segment tracheal stenosis in infancy. J Thorac Cardiovasc Surg 110:1504-1512, 1995 18. 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 19. Tsang V, Murday A, Gilbe C, et al: Slide tracheoplasty for congenital funnel-shaped stenosis. Ann Thorac Surg 48:632-635, 1989 20. Grillo HC: Slide tracheoplasty for long-segment congenital tracheal stenosis. Ann Thorac Surg 58:613-621, 1994 21. Cunningham MJ, Eavey RD, Vlahakes GJ, et al: Slide tracheoplasty for long-segment tracheal stenosis. Arch Otolaryngol Head Neck Surg 124:98-103, 1998 22. Lang FJW, Hurni M, Monnier P: Long-segment congenital tracheal stenosis: Treatment by slide tracheoplasty. J Pediatr Surg 34:1216-1222, 1999 23. Browdie DA, Bernstein RV, Johnson R: Materials for tracheoplasty: Which work? Which are best? J Thorac Cardiovasc Surg 113:810, 1997 (letter) 24. 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 112:1549-1560, 1996 25. Stokes JR, Heatley DG, Lusk RP, et al: The ‘‘Bridging Bronchus’’: Successful diagnosis and repair in a Stridulous Infant. Arch Otolaryngol Head Neck Surg 123:1344-1347, 1997
Discussion R. Filler (Toronto, Ontario): Certainly in terms of tissue, that kind of graft is a lot easier to sew in, say, than cartilage, and it has more stability to it than does a piece of pericardium that has been used by a lot of people. I guess it is still not enough tissue to reach the tops of the grafts in the very long stenoses.
You might like to say a word about the slide tracheoplasty. I have not used it, but there have been several reports in literature about sliding 2 pieces of trachea along each other and making a lumen out of it. And that certainly looks like an interesting concept that, as best I can read, the results are about the same as this.
TRACHEAL AUTOGRAFT: INTERMEDIATE RESULTS
One of the things I would like to point out in your paper was the use of cardiopulmonary bypass, which, I think, is very important. And I know a lot of people do not use cardiopulmonary bypass for this. I would just ask really one question, and that is the fate of this graft. Do you have any experimental evidence, or even in perhaps the autopsy findings of that child who died to know what happens to that graft? Does the cartilage remain and the rest of the tissue disappear? What is left in terms of structure? C.L. Backer (response): To address first the fate of the autograft and the child that did die, we did perform an autopsy, and the autograft had healed completely and looked essentially like normal trachea. Also, Dr Dodge-Khatami and I made an experimental model of this autograft. We have done this in about 30 rabbits. And at 2 weeks, 1 month, and 2 months, at the time of death the autograft looks vascularized and looks really indistinguishable from the other trachea. So for some reason, this thing seems to get a blood supply very quickly. We have not killed them as early as a day or 2 after the implant. But at least at 2 weeks after the surgery,
819
it looks like normal trachea, and it looks like it has a normal blood supply. We are investigating why that happens. With regard to the slide tracheoplasty, we have done this in 2 patients. One did quite well. The other child had a lot of trouble with granulation tissue. The slide tends to look like a figure of 8 when you get done. The spring of the cartilage sort of brings it in. And maybe we should have resected more of the sides of the cartilage. But that particular child then had tracheomalacia from that spring, required a stent, and ended up dying of granulation tissue and tracheal stenosis. I also find that the autograft operation is technically easier to do than the slide. When you do the slide you have to predetermine the midportion of the trachea when you cut it, which is not always easy to do, even with intraoperative bronchoscopy. The other thing about the slide is that if there is a posterior row of sutures at the upper trachea that you are operating under a shelf when you put those sutures in and tie them, and I found those somewhat difficult to do. With the autograft, everything is anterior and relatively straightforward.
JUNE 2000
Intermediate-Term Results of the Free Tracheal Autograft for Long Segment Congenital Tracheal Stenosis By Carl L. Backer, Constantine Mavroudis, Michael E. Dunham, and Lauren Holinger Chicago, Illinois
Background/Purpose: Since 1996 a new procedure—the free tracheal autograft—has been used to repair long segment congenital tracheal stenosis (LSCTS). The purpose of this report is to examine the intermediate-term results of that technique. Methods: Between January 1996 and July 1999, 10 infants underwent repair of LSCTS using a free tracheal autograft. Age ranged from 10 days to 23 months (mean age, 6.6 months). Six infants had a pulmonary artery (PA) sling; 5 had intracardiac anomalies. On cardiopulmonary bypass (CPB) the trachea was incised anteriorly through the area of stenosis. The midportion of the stenotic trachea was excised, and an end-to-end anastomosis was made posteriorly. The excised tracheal segment was trimmed and sutured in place anteriorly as a free autograft. In 5 patients the autograft was not long enough, and the upper portion of the tracheal opening was patched with pericardium.
required extracorporeal membrane oxygenation postoperatively for cardiac failure. The other 9 children are alive and well at 2 to 44 months postoperatively (mean follow-up, 24 months). One child had autograft dehiscence and required replacement of the autograft with an aortic homograft. Two children have tracheostomies at 6 and 36 months postoperatively. All children have had serial postoperative bronchoscopic examinations. Most recent bronchoscopies have shown widely patent tracheal lumina in all survivors.
Conclusion: Intermediate-term follow-up of children with a free tracheal autograft continues to support our use of this technique as our procedure of choice for infants with LSCTS. J Pediatr Surg 35:813-819. Copyright 娀 2000 by W.B. Saunders Company.
Results: There was 1 death 26 days postoperatively in a child that had simultaneous repair of tetralogy of Fallot and
INDEX WORDS: Tracheal stenosis, complete tracheal rings, tracheal autograft, tracheoplasty.
T
children that have undergone tracheal autograft for long segment congenital tracheal stenosis.
HE FIRST SUCCESSFUL repair of long segment congenital tracheal stenosis (LSCTS) at Children’s Memorial Hospital was with an anterior pericardial patch placed through a median sternotomy with the use of cardiopulmonary bypass in 1982.1 Over the next 14 years this procedure was used in 28 infants with a low operative mortality rate (7%) but a significant late mortality rate from airway complications (11%) and a relatively high reoperation rate (21%) for residual tracheal stenosis.2 In addition, these patients were noted to have frequent problems with granulation tissue requiring multiple bronchoscopies and tracheomalacia sometimes requiring prolonged intubation or stent placement.3,4 The tracheal autograft technique was used first essentially in desperation in an infant with an extremely dimunitive distal trachea, a pulmonary artery sling, and large ventricular septal defect (VSD).5 This report reviews our intermediate experience with infants and Journal of Pediatric Surgery, Vol 35, No 6 (June), 2000: pp 813-819
MATERIALS AND METHODS Between January 1996 and July 1999, 10 infants underwent repair of LSCTS using a free tracheal autograft technique. There were 7 boys and 3 girls. Age at surgery ranged from 10 days to 23 months (mean age, 6.6 months). Weight at surgery ranged from 2.9 to 10 kg (mean weight, 5.7 kg). All patients had cardiac echocardiography (ECHO), and most had a
From the Divisions of Cardiovascular-Thoracic Surgery and Otolaryngology, Northwestern University Medical School, and the Department of Surgery, The Children’s Memorial Hospital, Chicago, IL. Presented at the 1999 Annual Meeting of the American Academy of Pediatrics, Washington DC, October 8-10, 1999. Address reprint requests to Carl L. Backer, MD, Children’s Memorial Hospital, 2300 Children’s Plaza, m/c 22, Chicago, IL 60614. Copyright 娀 2000 by W.B. Saunders Company 0022-3468/00/3506-0001$03.00/0 doi:10.1053/js.2000.6847 813
814
BACKER ET AL
computed tomography (CT) scan with contrast. Six patients had a pulmonary artery (PA) sling. Five had intracardiac anomalies including tetralogy of Fallot (TOF, n ⫽ 2), ventricular septal defect (VSD), complete atrioventricular canal defect (CAVC), and atrial septal defect (ASD, n ⫽ 3). The number of complete tracheal rings ranged from 6 to 22 (mean, 15 rings). Five patients were transferred intubated and ventilated and required an urgent or emergent procedure. Seven of 10 patients were out-of-state referrals. Other associated congenital anomalies included patent ductus arteriosus (PDA, n ⫽ 5), tracheal right upper lobe (RUL) bronchus (n ⫽ 3), left superior vena cava (SVC, n ⫽ 2), subglottic stenosis (n ⫽ 2), hypoplastic right lung (n ⫽ 2), Down’s syndrome (n ⫽ 2), and absent left lung (n ⫽ 1). These patient characteristics are summarized in Table 1. All patients underwent rigid bronchoscopy within 2 days before the autograft or immediately after anesthetic induction for the autograft procedure. In most cases the bronchoscope itself could not be passed through the tracheal rings. In 3 cases not even the fine telescope (outer diameter, 2.5 mm) could be passed completely through the stenosis. Surgical approach was with a median sternotomy extended vertically onto the neck. Cardiopulmonary bypass was used in all cases to provide safe respiratory support while the trachea is being repaired. Bicaval cannulation (28°C) was used for cases in which an intracardiac anomaly was to be repaired, and right atrial cannulation (32°C) for all others. Intracardiac procedures were performed first, followed by PA sling repair, followed by tracheal repair. Intracardiac procedures in 5 patients included complete repair of VSD, ASD (n ⫽ 3), TOF (n ⫽ 2), and CAVC. In 6 patients, PA sling repair was performed. In 5 patients the left pulmonary artery (LPA) was transsected, moved anterior to the trachea, and reimplanted into the main pulmonary artery (MPA).6 In one patient with a hypoplastic right lung and a tiny right PA the left PA was translocated anterior to the trachea after tracheal division as originally described by Jonas et al.7 After intracardiac or PA sling repair, the entire trachea was extensively mobilized. This extended superiorly to the cricoid, and in most cases the isthmus of the thyroid gland was divided. Inferiorly, the dissection was carried below the carina and well onto the right and left mainstem bronchi. Bilateral hilar release and freeing of the PA from the pericardium also was included.8 The anterior trachea was incised with an 11 blade through the extent of the complete tracheal rings (Fig 1). An assessment was made as to how much trachea could be resected without causing excessive tension at the anastomosis. This excised midportion
Fig 1. Trachea with complete tracheal rings from ring 6 to just above the carina. Dotted line shows anterior incision to be made in trachea through the area of stenosis.
of the trachea is used for the free tracheal autograft (Fig 2). The 2 remaining ends of trachea are anastomosed posteriorly with multiple interrupted 6-0 PDS sutures (Ethicon Inc, Somerville, NJ) (Fig 3). These sutures are placed to keep the sutures and knots out of the tracheal lumen. The free tracheal autograft is used to patch the trachea anteriorly (Fig 3). This is accomplished by trimming the corners of the autograft and anchoring it in place with interrupted 6-0 PDS sutures (Fig 3). When the autograft can augment only the inferior portion of the anterior tracheal opening, the upper aspect of the tracheal incision is patched
Table 1. Patient Clinical Characteristics Patient No.
Age (mo)
Weight (kg)
1
2
4.5
2 3
5 2
4.9 4.2
4
0.3
2.9
5
12
10
6
8
6.2
7
6
5.0
8
6
6.0
9 23 9.9 10 2 3.5 Mean 6.6 ⫾ 6.7 5.7 ⫾ 2.4
Associated Anomalies
PA sling, VSD, LSVC, ASD, tracheal RUL Subglottic stenosis PA sling, PDA, hypoplastic right lung PA sling, PDA, LSVC, subglottic stenosis PA sling, hypoplastic right lung CAVC, PDA, Down’s syndrome TOF S/P shunt, S/P outflow patch, tracheal RUL PDA, Down’s syndrome, absent left lung PA sling, ASD PA sling, TOF, ASD, PDA
No. of Tracheal Rings
15 6 16 18 20 8 16 20 22 16 15.7 ⫾ 5.1
Fig 2. After the trachea has been opened anteriorly the midportion of the complete tracheal rings is excised to be used as the free tracheal autograft.
TRACHEAL AUTOGRAFT: INTERMEDIATE RESULTS
815
is stable, the paralysis is stopped, and the child is weaned from the ventilator over 3 to 5 days and extubated. Follow-up bronchoscopy is performed just before extubation, just before discharge, at 3 and 6 months, and then yearly.
RESULTS
Fig 3. The 2 ends of the trachea are anastomosed together posteriorly with interrupted sutures kept out of the lumen. The corners of the autograft are trimmed and the autograft is placed as an anterior patch with interrupted sutures.
with fresh autologous pericardium anchored with 6-0 PDS sutures (Fig 4). Pericardial augmentation was required in 5 patients. The mediastinum is copiously irrigated with an amikacin solution. Tissell glue (Baxter Healthcare Corp, Glendale, CA) is used to seal the suture lines on the trachea. Small hemoclips are placed in the soft tissue adjacent to the upper and lower aspects of the autograft for radiographic analysis of the position of the endotracheal (ET) tube. We try to keep the ET tube either (1) above the autograft if there is no pericardial augmentation or (2) in the midportion of the autograft in patients that have had pericardial augmentation. In patients in whom pericardial augmentation was used, the ET tube acts as a stent for the pericardial patch. No patient had a neck brace or chin stitch. The patient is kept paralyzed and ventilated for 3 to 5 days, at which time surveillance rigid bronchoscopy is performed. If the tracheal lumen is patent and the child
There was 1 early death (10%) and no late deaths. There was one significant complication—autograft dehiscence requiring replacement of the autograft with aortic homograft. Days of intubation ranged from 7 to 35 days (mean, 15 ⫾ 9.3 days), days of hospitalization ranged from 14 to 50 days (mean, 29.2 ⫾ 13.9 days). Most patients had 4 to 6 bronchoscopies during their hospitalization, although several ‘‘outliers’’ required multiple bronchoscopies. Currently, we are performing yearly bronchoscopies on all patients. Two patients still require tracheostomies at 6 and 36 months postrepair. Follow-up is complete in all patients and ranges from 2 to 44 months (mean, 24.1 ⫾ 16.4 months). Surgical findings, operative techniques, and clinical outcome are summarized in Table 2. The 1 early death occurred 26 days postoperatively in patient 7 in the series. This 6-month-old child had multiple congenital anomalies including tetralogy of Fallot, anomalous left coronary artery, agenesis of the left kidney, and extra digits. He had had 2 prior sternotomies for palliation of TOF with right ventricular outflow tract patch and modified Blalock-Taussig shunt. He underwent shunt takedown, complete repair of TOF, and autograft repair of 16 complete tracheal rings. He had progressive cardiac failure postoperatively and was placed on extracoporeal membrane oxygenation (ECMO) 12 hours after complete repair. He was supported on ECMO for 6 days. He became septic, and anasarca developed over a 2-week period. He then had multisystem organ failure and died 26 days postoperatively. Autopsy findings showed mulTable 2. Surgical Technique and Clinical Outcome Autograft Patient Length (cm) No.
1 2 3 4 5 6 7
Fig 4. For the infant with stenosis from just below the cricoid to the carina, the autograft is again taken from the midportion of the trachea. The autograft is placed inferiorly at the critical area for healing just above the carina. Note the autograft is trimmed preserving the upper corners. Pericardium anchored with interrupted sutures is placed superiorly to complete the tracheal reconstruction.
1.5 1.3 1.3 2.2 2.0 1.8 1.9
Pericardial Patch Length (cm)
Intubation (d)
Discharge (d)
Follow-Up (mo)
2.5 — 3.0 1.5 3.0 — —
13 20 44 7 14 38 7 19* 36 21 34 34 15 28 29 14 21 26 Died postoperative — — day 26 8 2.5 2.0 28 (Trach) 50 7 9 2.5 — 8 18 4 10† 2.0 — 35 49 2 Mean 1.9 ⫾ 0.4 2.4 ⫾ 0.6 15 ⫾ 9.3 29.2 ⫾ 13.9 24.1 ⫾ 16.4 *Required multiple readmissions over a 5-month period, eventually had a tracheostomy. †Had autograft dehisence and required emergent replacement of the autograft with an aortic homograft and later Palmaz stent.
816
BACKER ET AL
tiple areas of myocardial infarction and necrotizing bilateral pneumonia. The autograft repair was intact and without stenosis or evidence of infection. The child with autograft dehiscence underwent complete repair of PA sling, TOF, ASD, and autograft at 2 months of age. This child had a tracheal RUL bronchus and a stenotic ‘‘bridging bronchus.’’9 The autograft was divided in half longitudinally, and the 2 segments were both used for the anterior patch in an effort to not use pericardium. He had copious thick secretions that cultured positive for pseudomonas. On day 4 postoperation, during bronchoscopy, the autograft disrupted, and he had a massive air leak followed by desaturation and cardiac arrest. His chest was opened and the distal trachea was intubated directly. The upper autograft had lysed, and only the lower autograft could be identified. The patient was placed back on cardiopulmonary bypass (CPB) and the remaining autograft was removed. A cryopreserved aortic homograft was thawed and trimmed and placed as an anterior tracheal patch as described by Chahine et al.10 A vascularized muscle flap—the right pectoralis—was brought through an intercostal window to treat the pseudomonas mediastinitis.11 Because of collapse of the aortic homograft patch, a balloon expandable metallic (Palmaz) stent (20 mm length, 5 mm diameter) was placed 27 days after the reoperation.12 He was extubated 1 week later and discharged from the hospital 2 weeks after extubation. All patients currently are asymptomatic from a respiratory standpoint except for the 2 with tracheostomies. All have patent airways that on serial bronchoscopic examination appear to be growing normally with the patient. The autograft is difficult to differentiate from the surrounding trachea. The posterior end-to-end anastomosis has healed and has not been a source of restenosis. In 3 patients the pericardial portion of the repair has healed and presumably been replaced by ciliated pseudostratified columnar epithelium.13 In 2 other patients the pericardium has been the source of ‘‘patch tracheomalacia’’; and those 2 children have tracheostomies for that reason. DISCUSSION
We have reviewed our intermediate-term results with the technique of free tracheal autograft in 10 infants with long-segment congenital tracheal stenosis. The technique utilizes a partial tracheal resection, posterior end-to-end anastomosis, and anterior patching of the trachea with the resected piece—the tracheal autograft.5 Six of these patients had simultaneous repair of PA sling, and 5 had simultaneous repair of an intracardiac anomaly. Cardiopulmonary bypass was used in all patients. This greatly facilitated the tracheal diagnosis and repair. The telescope can be used in the trachea without ventilating the patient.
There is no risk of intraoperative asphyxia even with the tiniest tracheas (1.5 to 2.0 mm). The patency rate of the left pulmonary artery is greatly improved in pulmonary artery sling patients by using cardiopulmonary bypass.6 There were no complications secondary to cardiopulmonary bypass in this series. At a mean follow up of 24 months, 7 of 9 patients (78%) are essentially asymptomatic from a respiratory standpoint, and 2 are doing well with tracheostomies. We also have studied this technique in a laboratory model and have performed the tracheal autograft successfully in over 30 rabbits.14 The advantages of the autograft technique are several. The trachea in infants with LSCTS often is excessively long; hence, autograft ‘‘material’’ is readily available. There have been no complications related to the end-toend anastomosis and posterior suture line in this series. The autograft is already lined with respiratory epithelium. The cartilage intrinsically maintains the tracheal contour. The autograft has a potential for growth. In our laboratory model the autograft appears microscopically viable as early as 2 weeks after the repair.14 There seems to be minimal granulation tissue associated with the autograft especially at the critical junction of the patch with the carina. Our intermediate follow-up seems to indicate there is normal growth of the autograft. Other surgical alternatives for LSCTS include pericardial tracheoplasty,1,3,15 cartilage tracheoplasty,16-18 slide tracheoplasty,4,19,20-22 pulmonary homograft,23 and tracheal homograft.24 A summary of the number of patients treated with these techniques and the reported mortality rates are shown in Table 3. At our institution we used the pericardial tracheoplasty technique almost exclusively until 1996 (Fig 5). Although our operative mortality rate was low (7%), there was a significant late mortality from airway complications (11%) and a relatively high reoperation rate (21%) for residual stenosis.2 These patients required prolonged stenting of the trachea with the endotracheal tube and often required multiple bronchoscopies for dilation and removal of granulation tissue.3 Bando et al15 reported 12 patients having pericardial tracheoplasty with a mortality rate of 17%. Although we have changed to the tracheal autograft technique, we still use pericardium for the very long stenoses when the autograft is not long enough to patch the entire tracheal opening. By using a relatively short segment of pericardium in the upper trachea (away from the carina), the complications related to the use of pericardium are considerably fewer. The pericardium becomes replaced by ciliated pseudostratified columnar epithelium.13,15 We have used cartilage tracheoplasty successfully in 3 of 4 patients undergoing revision of pericardial tracheoplasty.2 Jaquiss et al17 reported excellent results with cartilage tracheoplasty, 6 patients (mean age, 3 months) with no mortality. Kamata et al18 reported quite different
TRACHEAL AUTOGRAFT: INTERMEDIATE RESULTS
817
Table 3. Published Results—Tracheoplasty Series Study
Technique
Year
Mean Age (mo)
No. of Patients
Mortality Rate (%)
Bando et al15 Backer et al2 Tsugawa et al16 Jaquiss et al17 Kamata et al18 Tsang et al19 Grillo8 Dayan et al4 Cunningham et al21 Lang et al22 Browdie et al23 Jacobs et al24 Backer et al (present study)
Pericardial tracheoplasty Pericardial tracheoplasty Cartilage tracheoplasty Cartilage tracheoplasty Cartilage tracheoplasty Slide tracheoplasty Slide tracheoplasty Slide tracheoplasty Slide tracheoplasty Slide tracheoplasty PA homograft Tracheal homograft Autograft
1996 1997 1988 1995 1997 1989 1994 1997 1998 1997 1997 1996 1999
7 6 10 3 5 6 10.5 yr 2 14 21 4 36 7
12 28 4 6 11 2 4 2 3 2 2 10 10
2 (17) 5 (18) 1 (25) 0 5 (45) 1 (50) 0 1 (50) 0 0 0 3 (30)* 1 (10)
*Nine of 10 had prior tracheal operation.
results, 11 patients (mean age, 5 months) with a 45% mortality rate. We have found cartilage (from rib) somewhat difficult to carve and implant. The cartilage is quite rigid and more difficult to achieve an air-tight seal. We and others have had the experience in which the cartilage is ‘‘too thick’’ and seems to ‘‘fall into the lumen,’’ occluding the lumen. Stokes et al25 reported on a patient in whom initial cartilage tracheoplasty from a rib had to be redone with ear cartilage for this very reason. The slide tracheoplasty technique was first reported in 2 patients by Tsang et al in 1989.19 In the past 10 years, 9 more patients have been reported with an overall mortality of 18%.8,4,21,22 Our experience comparing the 2 techniques has been that the autograft technique is easier to perform technically, is applicable to all patients with LSCTS (even those with complete tracheal rings from
Fig 5. Pericardial tracheoplasty. The trachea is incised anteriorly through the extent of the complete tracheal rings and patched anteriorly with autologous pericardium.
cricoid to carina), and results in an architecturally more desirable tracheal lumen. The slide tracheoplasty requires precise identification of the midportion of the stenosis before the initial tracheal transsection. This is not always easy to do either bronchoscopically or from external examination. If the trachea is transected initially in the wrong site, the 2 portions of trachea will not correspond and there will be a residual stenosis. 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 tie. In contrast, with the autograft technique, all the suturing is anterior and analogous to what surgeons are used to with the pericardial patch and cartilage tracheoplasty. Air leaks along the posterior suture line are particularly difficult to deal with. In the report of Grillo20 he noted having to perform an anterior vertical tracheal incision after the repair to deal with a persistent posterior air leak in an ‘‘inaccessible spot.’’ When the slide tracheoplasty is completed, the spring of the tracheal rings tends to form a ‘‘figure of 8’’ contour in cross section. This ‘‘unnatural’’ architectural configuration is a set-up for forces leading to granulation tissue and tracheomalacia. One of our patients had significant granulation tissue form at this waist and had flattening of the trachea requiring an endobronchial (Palmaz) stent and later a tracheostomy.4 The use of pulmonary homograft as an anterior tracheoplasty for infants with LSCTS was reported first by Browdie et al in 1997.23 Chahine et al10 reported the use of aortic homograft in 3 children, 1 with a gunshot wound, 1 with bronchial dehiscence after lung transplant, and 1 with an acquired tracheoesophageal fistula. We used this technique as rescue therapy for a child with autograft dehiscence. Although the homograft sealed well and with a muscle flap survived mediastinitis, it remained ‘‘floppy,’’ and patch tracheomalacia required an endobronchial (Palmaz) stent. Finally, Jacobs et al24 recently have reported the use of
818
BACKER ET AL
tracheal homograft for patients with tracheal stenosis from trauma, intubation, malignancy, and LSCTS. Their report details the use of a tracheal homograft in 10 infants with LSCTS, 9 of whom had a prior tracheal repair of some sort. Their survival rate of 70% in this complex group is extraordinary. We considered this option in the child in our series that had autograft dehiscence, but did not have a tracheal homograft available. Long segment congenital tracheal stenosis from com-
plete tracheal rings in children is often a life-threatening problem that frequently is associated with PA sling or major intracardiac anomalies. We recommend repair of the tracheal stenosis with the tracheal autograft technique and simultaneous repair of associated PA sling and cardiac anomalies. Care of these children after surgery is very demanding and requires close cooperation between pediatric otolaryngology and thoracic surgery departments.
REFERENCES 1. Idriss FS, DeLeon SY, Ilbawi MN, et al: Tracheoplasty with pericardial patch for extensive tracheal stenosis in infants and children. J Thorac Cardiovasc Surg 88:527-535, 1984 2. Backer CL, Mavroudis C, Dunham ME, et al: Reoperation after pericardial patch tracheoplasty. J Pediatr Surg 32:1108-1112, 1997 3. Dunham ME, Holinger LD, Backer CL, et al: Management of severe congenital tracheal stenosis. Ann Otol Rhinol Laryngol 103:351356, 1994 4. Dayan SH, Dunham ME, Backer CL, et al: Slide tracheoplasty in the management of congenital tracheal stenosis. Ann Otol Rhin Laryngol 106:914-919, 1997 5. 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 6. 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 7. Jonas RA, Spevak PJ, McGill T, et al: Pulmonary artery sling: primary repair by tracheal resection in infancy. J Thorac Cardiovasc Surg 97:548-550, 1989 8. Grillo HC: Surgical approaches to the trachea. Surg Gynecol Obstet 129:347-352, 1969 9. Gonzalez-Crussi FL, Padilla M, Miller JK, et al: Bridging bronchus: A previously undescribed airway anomaly. Am J Dis Child 130:1015, 1976 10. 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 11. Backer CL, Pensler J, Mavroudis C, et al: Vascularized muscle flaps for life threatening mediastinal wounds in children. Ann Thorac Surg 57:797-802, 1994 12. 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 13. Cheng ATL, Backer CL, Holinger LD, et al: Histopathological changes following pericardial patch tracheoplasty. Arch Otolaryngol Head Neck Surg 123:1069-1072, 1997
14. 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 vol L: 146-147, 1999 15. Bando K, Turrentine MW, Sun K, et al: Anterior pericardial tracheoplasty for congenital tracheal stenosis: Intermediate to long-term outcomes. Ann Thorac Surg 62:981-989, 1996 16. 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 17. Jaquiss RDB, Lusk RP, Spray TL, et al: Repair of long-segment tracheal stenosis in infancy. J Thorac Cardiovasc Surg 110:1504-1512, 1995 18. 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 19. Tsang V, Murday A, Gilbe C, et al: Slide tracheoplasty for congenital funnel-shaped stenosis. Ann Thorac Surg 48:632-635, 1989 20. Grillo HC: Slide tracheoplasty for long-segment congenital tracheal stenosis. Ann Thorac Surg 58:613-621, 1994 21. Cunningham MJ, Eavey RD, Vlahakes GJ, et al: Slide tracheoplasty for long-segment tracheal stenosis. Arch Otolaryngol Head Neck Surg 124:98-103, 1998 22. Lang FJW, Hurni M, Monnier P: Long-segment congenital tracheal stenosis: Treatment by slide tracheoplasty. J Pediatr Surg 34:1216-1222, 1999 23. Browdie DA, Bernstein RV, Johnson R: Materials for tracheoplasty: Which work? Which are best? J Thorac Cardiovasc Surg 113:810, 1997 (letter) 24. 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 112:1549-1560, 1996 25. Stokes JR, Heatley DG, Lusk RP, et al: The ‘‘Bridging Bronchus’’: Successful diagnosis and repair in a Stridulous Infant. Arch Otolaryngol Head Neck Surg 123:1344-1347, 1997
Discussion R. Filler (Toronto, Ontario): Certainly in terms of tissue, that kind of graft is a lot easier to sew in, say, than cartilage, and it has more stability to it than does a piece of pericardium that has been used by a lot of people. I guess it is still not enough tissue to reach the tops of the grafts in the very long stenoses.
You might like to say a word about the slide tracheoplasty. I have not used it, but there have been several reports in literature about sliding 2 pieces of trachea along each other and making a lumen out of it. And that certainly looks like an interesting concept that, as best I can read, the results are about the same as this.
TRACHEAL AUTOGRAFT: INTERMEDIATE RESULTS
One of the things I would like to point out in your paper was the use of cardiopulmonary bypass, which, I think, is very important. And I know a lot of people do not use cardiopulmonary bypass for this. I would just ask really one question, and that is the fate of this graft. Do you have any experimental evidence, or even in perhaps the autopsy findings of that child who died to know what happens to that graft? Does the cartilage remain and the rest of the tissue disappear? What is left in terms of structure? C.L. Backer (response): To address first the fate of the autograft and the child that did die, we did perform an autopsy, and the autograft had healed completely and looked essentially like normal trachea. Also, Dr Dodge-Khatami and I made an experimental model of this autograft. We have done this in about 30 rabbits. And at 2 weeks, 1 month, and 2 months, at the time of death the autograft looks vascularized and looks really indistinguishable from the other trachea. So for some reason, this thing seems to get a blood supply very quickly. We have not killed them as early as a day or 2 after the implant. But at least at 2 weeks after the surgery,
819
it looks like normal trachea, and it looks like it has a normal blood supply. We are investigating why that happens. With regard to the slide tracheoplasty, we have done this in 2 patients. One did quite well. The other child had a lot of trouble with granulation tissue. The slide tends to look like a figure of 8 when you get done. The spring of the cartilage sort of brings it in. And maybe we should have resected more of the sides of the cartilage. But that particular child then had tracheomalacia from that spring, required a stent, and ended up dying of granulation tissue and tracheal stenosis. I also find that the autograft operation is technically easier to do than the slide. When you do the slide you have to predetermine the midportion of the trachea when you cut it, which is not always easy to do, even with intraoperative bronchoscopy. The other thing about the slide is that if there is a posterior row of sutures at the upper trachea that you are operating under a shelf when you put those sutures in and tie them, and I found those somewhat difficult to do. With the autograft, everything is anterior and relatively straightforward.