Surgical management of congenital tracheal stenosis associated with tracheal bronchus and congenital heart disease Free

Abstract

OBJECTIVES

Congenital tracheal stenosis (CTS) is variable in patients with tracheal bronchus and congenital heart disease (CHD). Tracheoplasty remains a high-risk surgical treatment.

METHODS

From January 2007 to December 2014, 24 CTS patients (10 males and 14 females; age 20.6 ± 13.6 months) with tracheal bronchus and CHD underwent one-stage surgical correction. Clinical features of all patients included dyspnoea, or recurrent pulmonary infections. There was long-segment CTS in 13 cases (54%), and 4 cases were associated with a bridging bronchus. Less than 50% of normal tracheal size was identified in 21 cases. Complete tracheal or bronchial rings were identified in all cases. Operative techniques included: tracheal end-to-end anastomosis in 11 cases and slide tracheoplasty in 13 cases, which included 11 cases of right upper lobe bronchus (RULB) opposite side-slide tracheoplasty.

RESULTS

There were 2 operative deaths, due to postoperative tracheomalacia or residual main bronchial stenosis. The duration of postoperative hospital stay was 7–59 days, with an average of 19 days. Follow-up was completed in 19 cases. The follow-up duration was from 1 month to 7 years. Tracheal granulation occurred in 1 case. Clinical symptoms were significantly improved in the remaining patients.

CONCLUSIONS

Tracheal bronchi have special anatomical features. The techniques of tracheal end-to-end anastomosis or side-slide tracheoplasty can be used to correct tracheal stenosis with satisfactory outcomes.

INTRODUCTION

According to the recent literature, the spectrum of ‘tracheal bronchus’ encompasses a variety of bronchial anomalies. These malformations originate from the trachea or main bronchus, and are directed towards the upper lobe region [ 1 ]. When the entire right upper lobe bronchus (RULB) is displaced on the trachea, it is also called a ‘pig bronchus’ [ 2 ], which is common in some large mammals such as pigs, sheep and cattle.

The surgical treatment of congenital tracheal stenosis (CTS) remains a high-risk surgical procedure. Over the past 8 years, we have performed one-stage surgical correction of CTS complicated by congenital heart disease (CHD) in 97 patients, including 24 CTS patients associated with tracheal bronchus, and have obtained satisfactory operative outcomes.

MATERIALS AND METHODS

Approval for this retrospective study was obtained from the Shanghai Children's Medical Center and the affiliated Medical College of Jiao Tong University of Medicine Human Investigation Committee.

Patients and clinical presentation

From January 2007 to December 2014, one-stage surgical correction of CTS associated with tracheal bronchus and CHD was performed on 24 patients (10 males and 14 females). The displaced RULB originated directly from the trachea or the carina in these patients. The patients' age at the time of surgery ranged from 2 to 54 months (median, 20.6 ± 13.6 months). The patients' weight ranged from 4.0 to 15.5 kg (median, 9.5 ± 3.4 kg). Seven patients were less than a year old, and 8 patients were between 12 and 24 months old. Clinical features of all the patients showed paroxysmal, persistent dyspnoea or recurrent pulmonary infections. In addition to the patients' clinical features, their CHD and tracheal anomalies were diagnosed by echocardiographic examination and enhanced computed tomography (CT) images.

Based on enhanced 3D CT images, the CTS anatomy in 24 patients was characterized as one of the following morphologies:

• Main tracheal stenosis above the RULB (1 case), Fig. 1 A.

• Main tracheal stenosis between the RULB and the carina (10 cases), Fig. 1 B.

• Main tracheal stenosis associated with the carina trifurcation at the point of origin of the RULB from the carina (2 cases), Fig. 1 C.

• Long-segment main tracheal stenosis above the RULB to the carina (4 cases), Fig. 1 D.

• Tracheal bronchus associated with tracheal and main bronchus stenosis (3 cases), Fig. 1 E.

• Bridging bronchus, where the RULB originates from the carina, but the right middle and lower lobes are supplied by a bronchus that arises from the left main bronchus and crosses (bridges) the mediastinum from left to right (4 cases) [ 3 ], Fig. 1 F.

A fibre-optic bronchoscopy was performed preoperatively on all of the patients, and complete tracheal or bronchus rings were confirmed in all cases.

A CT scan was used to diagnose CTS. The transition zone above and below the stenotic segment was noted, and tracheal dimension data related to age and gender, as reported by Griscom and Wohl [ 4 ], were analysed. Freitag's classification system [ 5 ] was used to define the central airway stenosis. The degree of stenosis was assessed by a numerical code that is applicable to any site. Code 0 was assigned to non-appreciable stenosis and Codes 1, 2, 3 and 4 were assigned, respectively, to a 25, 50, 75 and 90% decrease in the cross-sectional area. Code 5 was assigned to complete obstruction. 75% of the patients in the group belonged to Code 3. If a stenotic segment was more than one-third of the tracheal length, it was defined as long-segment CTS. Otherwise, it was defined as local or short-segment CTS (Table 1 ). In addition, all of the patients in this cohort experienced CHD (Table 2 ).

Surgical management

All of the associated cardiovascular anomalies and tracheal repair were performed at the same stage through a median sternotomy, utilizing cardiopulmonary bypass (CPB) support. All cardiac defects were repaired, with the exception of 1 patient with a double outlet right ventricle and remote ventricular septal defect, who underwent a pulmonary artery (PA) banding operation. The cardiovascular anomalies were repaired before the tracheoplasty.

According to the tracheoplasty procedure, the patients were divided into two groups: Group A, tracheal end-to-end anastomosis ( n = 11), and Group B, slide tracheoplasty ( n = 13).

Tracheal end-to-end anastomosis

Eleven cases in Group A underwent end-to-end anastomosis, including main tracheal stenosis between the RULB and the carina in 8 cases; main tracheal stenosis above the RULB in 1 case and the involvement of main bronchus stenosis in 2 cases. For patients with stenosis between the RULB and the carina, the midpoint between the RULB and the carina was transected. The right side of the upper stump was then cut open to the orifice of the RULB. The carina was then laterally cut towards the side of the left or the right main bronchus to enlarge the orifices, especially when associated with left or right main bronchus stenosis. The upper stump was anastomosed to the lower part with continuous 5-0 polydioxanone (PDS; Ethicon) sutures to form a new carina.

Side-slide tracheoplasty

Thirteen cases in Group B with long-segment CTS received slide tracheoplasty, including 11 cases with side-slide tracheoplasty. For the patients with side-slide tracheoplasty, the opposite sides of each tracheal segment were opened longitudinally to points just beyond the extent of the stenosis, typically down to the carina distally. Since the orifice of the RULB was located on the right side of the proximal or distal segments of the trachea, longitudinal incisions were made on the left side of the tracheal segment opposite the side of the orifice of the RULB and on the right side of another tracheal segment, rather than using conventional anterior and posterior incisions. This procedure, introduced in this study, will be referred to as side-slide tracheoplasty. A sliding oblique anastomosis was performed with running 5-0 PDS suture (Ethicon) without compromising the orifice of the RULB. This is particularly beneficial in patients with severe main tracheal stenosis (Fig. 2 A–D). The posterior part of the anastomosis was performed first, as it was more difficult to expose. The anastomosis was carried inferiorly around the carina first. In line with our experience in this practice, we created an everting suture line along the entire length. This was best achieved by suturing from the outside of the lumen to the lining edge. After anastomosis, the suture line could hardly be seen in the lumen. Standard anterior/posterior incisions for the slide tracheoplasty were performed in 2 cases in Group B.

A fibre-optic bronchoscopy was repeated in the intensive care unit (ICU) to assess the tracheal morphology and to clear the airway secretions. Weaning from mechanical ventilation was initiated after stabilization, and the patients were treated with glucocorticoid atomization (Budesonide, AstraZeneca Pty Ltd, Australia), which can reduce the airway inflammatory response. Budesonide inhalation, 200–400 µg, was administered daily, divided into 2–4 doses per day.

RESULTS

The cardiac anomalies were repaired successfully in all patients with the exception of 1 PA banding patient. The CPB time ranged 51–195 min (median, 82 ± 45.61 min). The aortic clamp time ranged 0–61 min (median, 47 ± 27.19 min). Owing to more complicated tracheal morphology and severe symptoms preoperatively, the patients in Group B spent more time in the ICU and hospital than those in Group A. No statistical difference in mortality was identified between the two groups during hospitalization. The average hospital stay was 19 days after surgery (7–59 days). The clinical operative results are summarized in Table 3 .

There were 2 hospital deaths, 1 in Group A and 1 in Group B, neither of which was related to cardiovascular disease. One of these was a 2-month old female patient with severe respiratory distress preoperatively. The CT scan showed a PA sling, and a bridging bronchus associated with main tracheal and intermediate bronchial stenosis. The patient underwent an emergency operation. After CPB was established and the PA sling was repaired, a 2.8-mm fibre bronchoscope could not pass through the most narrow segment. The midpoint of the narrow segment was then transected. Severe intimal oedema and tracheomalacia with complete tracheal rings were identified. The narrowed tracheal lumen was found to be less than 2 mm. After side-slide tracheoplasty, the bronchoscope could pass through the tracheal anastomosis, but we were still unable to intubate the patient with a 3.5-mm tracheal tube. Owing to severe tracheomalacia, the tracheal wall developed paradoxical motion with respiratory inspiration (dynamic collapse). High airway pressure produced no improvement. The patient's family refused extracorporeal membrane oxygenation; and the patient died from respiratory failure 3 days postoperatively.

The other patient was a 6-month old male with a PA sling and tracheal bronchus associated with tracheal and main bronchus stenosis. A fibre-optic bronchoscopy showed complete tracheal rings and bronchus rings. The patient could not be weaned from the ventilator, and residual bronchus rings existed after tracheal end-to-end anastomosis. The patient died from a pulmonary infection 20 days postoperatively.

Postoperative follow-up was completed on 19 patients, ranging from 1 month to 7 years (median, 3.1 years), and clinical symptoms were significantly improved in 18 cases. Postoperative CT scans showed adequate tracheal size and morphology (Fig. 3 A and B). One patient who had tetralogy of Fallot associated with bridging bronchus and long-segment CTS subsequently died. The patient underwent side-slide tracheoplasty in the early phase. He was weaned from the ventilator 3 days postoperatively and discharged from hospital without incident. Two months later, the patient developed respiratory distress and pneumonia. A bronchoscopic examination showed granulation around the anastomosis edges, and the patient subsequently died from respiratory failure. Since then, we have used the everting suture technique in tracheoplasty, and the intraoperative bronchoscopic visualization has shown a satisfactory luminal area without the ridge of cartilage inserted into the lumen. A follow-up bronchoscopy at 1 month postoperatively showed a satisfactory appearance without any evidence of granulation tissue (Fig. 4 A–C).

DISCUSSION

It was reported in Ghaye's review that a tracheal bronchus was described by Sandifort in 1785 as an RULB originating from the trachea [ 2 ]. Tracheal bronchus in humans is frequently combined with CTS or CHD. The tracheal bronchi in this study included the RULBs originating from either the trachea or the carina. Four patients in the group had a bridging bronchus. The lower bifurcation is also called a pseudocarina [ 6 ], whose thoracic vertebral level is T5–6 [ 7 ]. The nosology of the bridging bronchus is contentious. We consider a bridging bronchus to be a type of congenital bronchial abnormality belonging to the tracheal bronchus spectrum. A bridging bronchus is often associated with long tracheal and intermediate bronchial stenosis. From the perspective of surgery, various tracheal bronchi, including ‘pig bronchus’, carina trifurcation and bridging bronchus, have the same operative indications.

Surgical management is indicated for children with significant respiratory symptoms. It is generally believed that paediatric patients can tolerate a narrowing of the trachea of up to 50%. Tracheal narrowing of less than 50% of the normal tracheal diameter (87.5% of the patients in this review) or long-segment CTS (involvement of the bronchi) is a crucial factor in determining the necessity for surgery. For patients with mild (>60% of the normal tracheal diameter) CTS, we generally only operate on the cardiovascular anomalies. For patients with CTS between 40 and 50%, the surgical option is determined by clinical symptoms. Even for patients with mild CTS (3 patients in this review), and exhibiting complete tracheal rings identified by fibre bronchoscopy and frequent wheezing, we are willing to consider the surgical option.

Eleven cases underwent end-to-end tracheoplasty in Group A. Overall, their stenotic segments were less than one-third of the tracheal length. The main advantages of end-to-end tracheoplasty are the relatively uncomplicated surgical procedure and rapid healing. For patients with a stenotic segment of more than 30% of tracheal length or more than 1.5 cm, end-to-end tracheoplasty may result in high anastomotic tension. This can lead to anastomotic rupture or leaks, and scar tissue regeneration, especially in infants [ 8 ]. One patient with bilateral bronchus stenosis died from residual bronchus rings and pulmonary infection in the early postoperative period. Hence, for patients with main bronchus stenosis, the carina is cut open into the orifice of the main bronchus, and then end-to-end tracheoplasty is performed [ 9 ].

Slide tracheoplasty was first described in 1989 by Tsang [ 10 ], and was initially designed for long-segment CTS. Slide tracheoplasty has become the preferred technique in CTS treatment [ 11 , 12 ], and appears to be superior to other techniques [ 13 , 14 ]. We first performed slide tracheoplasty to correct CTS in 2001 [ 15 ], and have completed numerous cases in the last 3 years. At present, we have abandoned other tracheoplasty techniques, in favour of end-to-end tracheoplasty and slide tracheoplasty. For slide tracheoplasty, the cartilage corners should be trimmed from the top of each segment in order to avoid being inserted into the lumen, or the formation of granulation tissue. The formation of tracheal granulation tissue did occur in the early period of our work in 1 case. Since then, we have improved our techniques, including using an everting suture, trimming the cartilage corners, preserving the blood supply and using postoperative glucocorticoid atomization treatment. Although we have no evidence to prove that these procedures can prevent granulation formation around the site of the tracheal anastomosis, no obvious insertion of the tracheal cartilage ridge has been documented in the postoperative fibre-optic bronchoscopy since then. In particular, the use of the modified technique of the everting suture effectively limits the presence of redundant tissue in the lumen and should help reduce the risk of anastomotic complications. Some surgeons have experienced success with this technique in atrial anastomosis in human lung transplantation [ 16 ].

The original slide tracheoplasty procedure described by Tsang et al. [ 10 ], and used by the Cincinnati group [ 13 ], indicated that the superior aspect of the trachea was incised anteriorly and the inferior aspect was incised posteriorly. Grillo [ 17 ] proposed an alternative, which is also used by the Chicago group [ 18 ], which places the incision on the superior aspect of the trachea posteriorly and the inferior aspect of the trachea anteriorly, which can then be extended to the bronchus. In the patients in Group B, the orifices of the RULB were located on the right side of the main trachea. The slide tracheoplasty was modified for those patients with long-segment CTS. Longitudinal incisions were made on the left side of the tracheal segment opposite the side of the orifice of the RULB and on the right side of another tracheal segment, rather than using anterior and posterior incisions. We refer to this as side-slide tracheoplasty, which has the following advantages: (i) it is easier to expose and stitch than the conventional incision, especially the anastomosis on the posterior part due to the presence of the RULB; (ii) it avoids distorting the orifice of the RULB, especially in patients with severe main tracheal stenosis and tracheomalacia, as the suture lines are far from the orifice and (iii) one can extend the inferior incision to the bronchus, if associated with bronchial stenosis. Side-slide tracheoplasty necessitates lateral dissection to perform a lateral incision, which could risk damage to the lateral blood supply. However, the dissection would not damage the opposite side, which then preserves the blood supply on one side.

In conclusion, tracheal bronchi have special anatomical features. The techniques of tracheal end-to-end anastomosis or side-slide tracheoplasty can be used to correct tracheal stenosis with satisfactory outcomes. Side-slide tracheoplasty is a very satisfactory alternative in the surgical treatment of long-segment CTS with tracheal bronchus.

Funding

This work was supported by Shanghai Hygiene Science Research (20134340) and National Nature Science Fund of China (81370117).

Conflict of interest: none declared.

REFERENCES

Author notes