https://orcid.org/0000-0002-1858-1523
Abstract
Children with congenital tracheal stenosis born in the developing world face a high risk of mortality due to limited access to proper treatment. Patients who required preoperative respiratory support were suspected to have poor survival after slide tracheoplasty; however, this was not clearly demonstrated in the previous studies. This study aims to investigate the impact of preoperative respiratory conditions on outcomes of slide tracheoplasty.
From 2016 to 2022, children who underwent slide tracheoplasty were retrospectively reviewed. Patients with respiratory distress requiring emergency operations (group A) were compared with patients in stable condition who were scheduled for surgery (group B).
Perioperative results revealed that group A (n = 43) had a longer bypass time (P < 0.001), operation time (P = 0.01), postoperative ventilation time (P < 0.001) and length of intensive care unit stay (P = 0.00125) than group B (n = 60). The early mortality rate was 7.8%, and the actuarial 5-year survival rate was 85.3%. The cumulative incidence test revealed that group A was highly significant for overall mortality [sudistribution (SHR) 4.5; 95% confidence interval (CI) 1.23–16.4; P = 0.023]. Risk factors for overall mortality were prolonged postoperative ventilation time (hazard ratio 3.86; 95% CI 1.20–12.48; P = 0.024), bronchial stenosis (hazard ratio 5.77; 95% CI 1.72–19.31; P = 0.004), and preoperative tracheal mucositis (hazard ratio 5.67; 95% CI 1.51–21.31; P = 0.01). Four patients needed reintervention during a follow-up of 28.4 months (interquartile range 15.3–47.3).
Preoperative respiratory distress negatively affected the outcomes of patients who required slide tracheoplasty. Therefore, early detection of congenital tracheal stenosis and aggressive slide tracheoplasty are crucial and obligatory to enhance long-term survival in this lethal congenital airway disease.
INTRODUCTION
Congenital tracheal stenosis (CTS) with complete tracheal rings usually requires medical treatment; this life-threatening condition is not rare, especially among symptomatic small children [1]. Although frequent observation may be indicated in a few asymptomatic patients [2], most patients diagnosed with CTS will need surgical repair [3]. Congenital heart defects (CHDs) associated with CTS are common and sometimes result in a more complicated decision-making process for surgical repair [4, 5]. Slide tracheoplasty (STP) has become the procedure of choice for treating CTS in high-income countries, achieving good outcomes with survival rates ranging from 88.2% to 95% [6–9]. The multidisciplinary team approach has been applied to treat patients diagnosed with CTS and has become the worldwide standard in high-income countries due to the complexity of the procedure, high demand for resources and request for collaborations to improve the outcomes of these patients [10, 11].
However, the outcomes of patients who are diagnosed with CTS in the developing world are still limited, with an assumption of high mortality and morbidity due to resource constraints. Patients with CTS associated with CHD pose real challenges for underdeveloped healthcare systems, which face difficulty in decision-making given the limitations of the optimal treatment. Before 2016, despite many attempts to treat patients with a diagnosis of CTS in North Vietnam, almost all patients died regardless of whether medical or surgical treatment was selected. This airway lesion has become a matter of utmost concern for paediatricians and the families of patients, as it has among the highest fatality rates. The 1st STP was successfully performed in our hospital in 2016, with support from Samsung Medical Center doctors (Prof. T.G.J.), concomitant with total correction of tetralogy of Fallot. After this initial success, a team consisting of 1 paediatric cardiac surgeon (Truong Ly Thinh Nguyen), 1 pneumologist (Chuong Thanh Le) and 1 cardiac intensivist (Thuc Van Dang) was dispatched to the Shanghai Children’s Hospital to learn how to improve the management of patients with CTS in 2017. Since then, the congenital cardiac team has performed STPs regularly at the Vietnam National Children’s Hospital. However, many patients presented to our hospital with severe respiratory distress due to a lack of knowledge and experience in the early detection of this catastrophic congenital condition, as well as resource limits of preoperative respiratory care to maintain adequate ventilation.
This study aims to evaluate the impact of preoperative respiratory distress on outcomes for children who are diagnosed with complete tracheal rings, and also analyse the intermediate results of STP in treating CTS in a single centre situated in a resource-constrained country.
PATIENTS AND METHODS
Patients and definitions
The study was approved by the Ethics Committee of the Hospital Review Board (Ref. No. 801/BVNTW-HDDD; 4 May 2022). A retrospective review was performed for all data collected, and informed consent was waived according to our National Legislation and Institutional Requirements.
From September 2016 to December 2022, patients underwent surgical repair for CTS due to complete tracheal rings at the Vietnam National Children’s Hospital. Patients with CTS categorized as class II or III by the Anton-Pacheco classification who underwent STP were included in this research. Exclusion criteria included patients with complete tracheal rings who had tracheoplasty by other methods such as resection with end-to-end anastomosis, pericardial patch tracheoplasty, tracheoplasty with tracheal autograph and patients with tracheal stenosis acquired after prolonged mechanical ventilation. Patients admitted to the hospital with respiratory distress requiring noninvasive or mechanical ventilation support were indicated for and underwent delayed emergency STP (group A). Patients who had indications for STP but were in stable condition were referred for a scheduled operation (group B).
Bronchoscopy was indicated in all patients in our study due to a suspicion of CTS to confirm the diagnosis. However, due to resource constraints (our hospital only has 1 smallest available fibre-optic bronchoscope of 2.8 mm diameter and 1 of 3.2 mm; we do not have access to optical coherence tomography), bronchoscopy was performed only once before the operation, usually at the time of diagnosis, and not indicated as a routine procedure if the postoperative course had a normal progression. Preoperative computed tomography was usually performed in patients with CTS in our study to help visualize the anatomy of any tracheal and associated cardiac lesions, except for some patients with extremely severe preoperative conditions. The anatomy of the CTS was classified according to Speggiorin et al. [12]. Long-segment CTS was defined as a stenosis length consisting of more than 50% of the trachea [13]. The Anton-Pacheco classification was used to determine the severity of CTS as either mild, moderate or severe stenosis [14]. Echocardiography was performed to detect associated intracardiac lesions in all patients. Bronchial stenosis was defined intraoperatively as a complete bronchial ring in at least 1 bronchus, except the superior tracheal bronchus. When the lining of the trachea turned yellow or the mucosa began to peel off from the wall of the trachea in isolated areas or conjunction with the full length of the airway according to intraoperative visualization, the patient was considered to have tracheal airway mucositis. Prolonged mechanical ventilation was defined as the need for mechanical ventilation for more than 5 days. Early mortality was defined as in-hospital mortality or death within 30 days of the operation. Reintervention was defined as postoperative reoperation or bronchoscopy intervention for balloon dilatation or granulation removal.
For those who had associated cardiac lesions, the follow-up was completed in the outpatient clinic of the Cardiology Department of the Heart Center, but for those who had CTS without intracardiac lesions, the follow-up was completed at the Department of Pneumology. All patients had regular follow-up at our hospital at 1 month, 3 months, 6 months and every year after operation. During follow-up, bronchoscopy and computed tomography are indicated in symptomatic patients. Patients were contacted at the latest follow-up via their parent’s telephone number, and data were collected in the hospital database for patient information.
The early mortality and overall mortality; duration of postoperative mechanical ventilation; length of intensive care unit stay; anastomotic dehiscence need reoperation; recurrent tracheal stenosis required postoperative intervention; and other outcome variables were evaluated to assess the impact of preoperative condition on unfavourable outcomes.
Surgical techniques and postoperative care
Almost all operations in this study were performed by Truong Ly Thinh Nguyen, who applied the surgical techniques for STP developed by Shanghai Children’s Hospital with some modifications [5, 15]. Associated intracardiac defects were concomitantly repaired using cardiopulmonary bypass via a sternotomy approach before STP. The trachea was exposed and mobilized from the larynx to the main bronchus to identify the affected tracheal stenosis. The distal end of the CTS was always identified as the carina. The trachea was divided at the midpoint of the stenosis, and further mobilization of the anterior and posterior of the trachea was performed while avoiding circumferential dissection to preserve the lateral tissue blood supply. The thyroid isthmus is divided and the upper end of the trachea is freed along the anterior and posterior sides, avoiding dissection into the lateral side to protect the recurrent laryngeal nerve. Associated complete bronchial stenoses were treated with a switch incision from the carina to the bronchial stenosis, and STP was performed from the affected bronchus up to the carina continuously, similar to the regular surgical approach. The upper tracheal segment was incised anteriorly, and the lower tracheal segment was incised posteriorly and anastomosed by continuous polydioxanone sutures of 5.0 or 6.0. The suture ran from the inferior to the superior and then stopped 0.5 cm before the end of the incision (Fig. 1). We modified the technique as previously described by the Shanghai group using polydioxanone interrupted sutures to complete the remaining tracheal anastomosis so that the upper part of the reconstructed trachea may be used for tracheostomy if needed by simply removing the interrupted suture without interfering with the remaining tracheal anastomosis (Video 1).
(A) Congenital tracheal stenosis with anatomical landmark. (B) The trachea was divided at the middle point of the stenotic segment. (C) The upper part of the trachea was longitudinally open in the anterior wall, and the lower part was open in the posterior wall. (D) Two parts were slid together using a running suture. The upper part of the anastomosis was completed with interrupted sutures.
Slide tracheoplasty for long-segment congenital trachea stenosis (CTS).
All patients were kept in the intensive care unit under sedation and muscle relaxation for 12–24 h and then slowly weaned off ventilation with continuous EtCO2 monitoring. In cases needing reintervention for postoperative tracheal stenosis due to granuloma, balloon dilatation was applied via bronchoscopy.
Statistical analysis
Data were collected retrospectively from hospital medical records, including preoperative conditions, perioperative variables, postoperative course and follow-up data. Patients were stratified into 2 groups according to the preoperative respiratory condition, and a comparison was made between groups focusing on the anatomies, perioperative variables, overall mortality and morbidity during follow-up. Continuous variables are expressed as medians with interquartile ranges (IQRs) or mean (standard deviations) as appropriate, and comparisons between groups were performed using the Kruskal–Wallis test. Categorical variables are represented as frequencies and percentages and were compared using Pearson’s chi-squared test if the frequency was >5 and Fisher’s exact test otherwise. The risk factors for overall mortality were analysed using multivariate Cox proportional hazards regression analysis. Multivariable risk factors are reported as hazard ratios (HRs) and their 95% confidence intervals (CIs) with a value of P < 0.05. The cumulative incidence function curve was plotted with the Fine-Gray competing risk regression model comparing the hazards of reoperation between the 2 groups. The survival functions were estimated taking into account competing risk with Aalen Johansen methods. The Kaplan–Meier method was used to estimate survival. All statistical analyses were performed with the R software package (version R-4.2.1; RStudio 2022.02.01 Build 461).
RESULTS
From September 2016 to December 2022, 103 patients who underwent STP at the Vietnam National Children’s Hospital were retrospectively reviewed. There were 43 patients (41.7%, 43/103) in group A and 60 patients (58.3%, 60/103) in group B. Eighty-four patients (81.6%) had long-segment tracheal stenosis (>50% of the tracheal length) and 33 patients (32%) had excessive long-segment stenosis (>80% of the tracheal length). Seventy-two patients (69.9%) had congenital heart disease. Due to some patients’ extremely critical preoperative circumstances preventing them from being transported to the CT scan area, multislice computed tomography was accessible for 92 patients. The anatomy of the trachea included normal tracheobronchial arborization (64.1%, 59/92), right tracheal bronchi (10.9%, 10/92), tracheal trifurcation (22.8%, 21/92) and a single lung (2.2%, 2/92).
The median age and weight were 7 months (IQR 3.5–13.4 months) and 6.7 kg (IQR 5.3–9.0 kg), respectively. Eight patients (7.8%) were <2 months of age, 45 patients (43.7%) were <6 months of age and 75 patients (72.8%) were <1 year of age. Patients in group A were significantly smaller (P = 0.002) and younger (P = 0.009) than those in group B. Preoperative ECMO support was performed on 2 patients in group A.
Associated cardiovascular anomalies account in 72 patients (69.9%), with equal distribution (P = 0.39) of RACHS-1 stratification between group A (74.4%, 32/43) and group B (66.7%, 40/60). These cardiovascular anomalies simultaneously underwent surgical repair with cardiopulmonary bypass, followed by STP to complete total correction. Concomitant congenital malformations were present in 8 patients (7.8%): 4 patients had anorectal malformations, 1 had trisomy 21, 1 had hydrocephalus, 1 had a cleft palate and 1 had VACTERL/VATER. The patient demographics and perioperative data are shown in Table 1.
Perioperative variables between groups
| Variables | Overall (N = 103) | Group A (preoperative unstable condition) (n = 43) | Group B (preoperative stable condition) (n = 60) | P-value |
|---|---|---|---|---|
| Age at operation (months), median (IQR) | 7 (3.6–13.4) | 3.9 (2.3–6.3) | 10.2 (6.5–19.2) | 0.002 |
| Weight (kg), median (IQR) | 6.7 (5.3–8.95) | 5.5 (4.65–6.5) | 8 (5.9–9.8) | 0.009 |
| Age groups, n (%) | ||||
| ≤2 months of age | 8 (7.8) | 6 (14) | 2 (3.3) | 0.07 |
| >2 months of age | 95 (92.2) | 37 (86) | 58 (96.7) | |
| None CHD | 31 | 11 | 20 | |
| Associated cardiac lesions, n (%) | 72 (69.9) | 32 (74.4) | 40 (66.7) | 0.39 |
| Risk category 2 | 8 | 1 | 7 | |
| Risk category 3 | 63 | 30 | 33 | |
| Risk category 4 | 1 | 1 | 0 | |
| Positive preoperative bacterial culture in the respiratory track, n (%) | 17 (16.5) | 14 (32.6) | 3(5.0) | <0.001 |
| Continuous stridor, n (%) | 90 (87.3) | 43 (100) | 47 (78.3) | 0.005 |
| Length of tracheal stenosis, median (IQR) | 4.5 (3.5–5) | 4 (3.5–5) | 5.0 (3.7–5.1) | 0.03 |
| Tracheal airway mucositis, n (%) | 43 (41.7) | 31 (72.1) | 12 (20) | <0.001 |
| Tracheal anatomy on msCT (92 patients) | ||||
| Normal, n (%) | 59/92 (64.1) | 32/40 (80) | 27/52 (51.9) | |
| Tracheal right upper lobe bronchus, n (%) | 10 (10.9) | 2 (5) | 8 (15.4) | |
| CTS—Trifurcation, n (%) | 21 (2.8) | 5 (12.5) | 16 (30.8) | |
| CTS—Single lung, n (%) | 2 (2.2) | 1 (2.5) | 1 (1.9) | |
| Bronchoplasty, n (%) | 10 (9.7) | 4 (9.3) | 6 (10) | 1 (OR = 0.9) |
| Bypass time (minutes), median (IQR) | 110 (81.5–138) | 127 (105–172) | 98 (69.5–121) | <0.001 |
| ACC time (hours), median (IQR) | 0 | 0 (0–4) | 0 | 0.02 |
| Operation time (hours), median (IQR) | 4.5 (4–5) | 4.5 (4–5.9) | 4 (3.5–5) | 0.01 |
| Mechanical ventilation time (hours), median (IQR) | 96 (56–157) | 138 (96–224) | 72 (48–117) | <0.001 |
| Length of ICU stay (days), median (IQR) | 8 (5–14) | 12 (9–18.5) | 6 (5–9) | 0.001 |
| Length of hospital stay (days), median (IQR) | 20 (14–26) | 24 (16.5–36) | 16 (13–21.5) | 0.004 |
| Reoperation, n (%) | 6 (5.8) | 5 (11.6) | 1 (1.7) | 0.08 (OR = 0.13) |
| Sepsis, n (%) | 11 (10.7) | 9 (20.9) | 2 (3.3) | 0.007 (OR = 0.13) |
| Early mortality, n (%) | 8 (7.8) | 6 (14) | 2 (3.3) | 0.06 (OR = 0.21) |
| Overall mortality, n (%) | 15 (14.6) | 12 (27.9) | 3 (5.0) | 0.002 (OR = 0.13) |
| Variables | Overall (N = 103) | Group A (preoperative unstable condition) (n = 43) | Group B (preoperative stable condition) (n = 60) | P-value |
|---|---|---|---|---|
| Age at operation (months), median (IQR) | 7 (3.6–13.4) | 3.9 (2.3–6.3) | 10.2 (6.5–19.2) | 0.002 |
| Weight (kg), median (IQR) | 6.7 (5.3–8.95) | 5.5 (4.65–6.5) | 8 (5.9–9.8) | 0.009 |
| Age groups, n (%) | ||||
| ≤2 months of age | 8 (7.8) | 6 (14) | 2 (3.3) | 0.07 |
| >2 months of age | 95 (92.2) | 37 (86) | 58 (96.7) | |
| None CHD | 31 | 11 | 20 | |
| Associated cardiac lesions, n (%) | 72 (69.9) | 32 (74.4) | 40 (66.7) | 0.39 |
| Risk category 2 | 8 | 1 | 7 | |
| Risk category 3 | 63 | 30 | 33 | |
| Risk category 4 | 1 | 1 | 0 | |
| Positive preoperative bacterial culture in the respiratory track, n (%) | 17 (16.5) | 14 (32.6) | 3(5.0) | <0.001 |
| Continuous stridor, n (%) | 90 (87.3) | 43 (100) | 47 (78.3) | 0.005 |
| Length of tracheal stenosis, median (IQR) | 4.5 (3.5–5) | 4 (3.5–5) | 5.0 (3.7–5.1) | 0.03 |
| Tracheal airway mucositis, n (%) | 43 (41.7) | 31 (72.1) | 12 (20) | <0.001 |
| Tracheal anatomy on msCT (92 patients) | ||||
| Normal, n (%) | 59/92 (64.1) | 32/40 (80) | 27/52 (51.9) | |
| Tracheal right upper lobe bronchus, n (%) | 10 (10.9) | 2 (5) | 8 (15.4) | |
| CTS—Trifurcation, n (%) | 21 (2.8) | 5 (12.5) | 16 (30.8) | |
| CTS—Single lung, n (%) | 2 (2.2) | 1 (2.5) | 1 (1.9) | |
| Bronchoplasty, n (%) | 10 (9.7) | 4 (9.3) | 6 (10) | 1 (OR = 0.9) |
| Bypass time (minutes), median (IQR) | 110 (81.5–138) | 127 (105–172) | 98 (69.5–121) | <0.001 |
| ACC time (hours), median (IQR) | 0 | 0 (0–4) | 0 | 0.02 |
| Operation time (hours), median (IQR) | 4.5 (4–5) | 4.5 (4–5.9) | 4 (3.5–5) | 0.01 |
| Mechanical ventilation time (hours), median (IQR) | 96 (56–157) | 138 (96–224) | 72 (48–117) | <0.001 |
| Length of ICU stay (days), median (IQR) | 8 (5–14) | 12 (9–18.5) | 6 (5–9) | 0.001 |
| Length of hospital stay (days), median (IQR) | 20 (14–26) | 24 (16.5–36) | 16 (13–21.5) | 0.004 |
| Reoperation, n (%) | 6 (5.8) | 5 (11.6) | 1 (1.7) | 0.08 (OR = 0.13) |
| Sepsis, n (%) | 11 (10.7) | 9 (20.9) | 2 (3.3) | 0.007 (OR = 0.13) |
| Early mortality, n (%) | 8 (7.8) | 6 (14) | 2 (3.3) | 0.06 (OR = 0.21) |
| Overall mortality, n (%) | 15 (14.6) | 12 (27.9) | 3 (5.0) | 0.002 (OR = 0.13) |
ACC time: aortic cross-clamp time; CHD: congenital heart disease; CTS: congenital tracheal stenosis; ICU: intensive care unit; IQR: interquartile range; msCT: multislice computed tomography.
Perioperative variables between groups
| Variables | Overall (N = 103) | Group A (preoperative unstable condition) (n = 43) | Group B (preoperative stable condition) (n = 60) | P-value |
|---|---|---|---|---|
| Age at operation (months), median (IQR) | 7 (3.6–13.4) | 3.9 (2.3–6.3) | 10.2 (6.5–19.2) | 0.002 |
| Weight (kg), median (IQR) | 6.7 (5.3–8.95) | 5.5 (4.65–6.5) | 8 (5.9–9.8) | 0.009 |
| Age groups, n (%) | ||||
| ≤2 months of age | 8 (7.8) | 6 (14) | 2 (3.3) | 0.07 |
| >2 months of age | 95 (92.2) | 37 (86) | 58 (96.7) | |
| None CHD | 31 | 11 | 20 | |
| Associated cardiac lesions, n (%) | 72 (69.9) | 32 (74.4) | 40 (66.7) | 0.39 |
| Risk category 2 | 8 | 1 | 7 | |
| Risk category 3 | 63 | 30 | 33 | |
| Risk category 4 | 1 | 1 | 0 | |
| Positive preoperative bacterial culture in the respiratory track, n (%) | 17 (16.5) | 14 (32.6) | 3(5.0) | <0.001 |
| Continuous stridor, n (%) | 90 (87.3) | 43 (100) | 47 (78.3) | 0.005 |
| Length of tracheal stenosis, median (IQR) | 4.5 (3.5–5) | 4 (3.5–5) | 5.0 (3.7–5.1) | 0.03 |
| Tracheal airway mucositis, n (%) | 43 (41.7) | 31 (72.1) | 12 (20) | <0.001 |
| Tracheal anatomy on msCT (92 patients) | ||||
| Normal, n (%) | 59/92 (64.1) | 32/40 (80) | 27/52 (51.9) | |
| Tracheal right upper lobe bronchus, n (%) | 10 (10.9) | 2 (5) | 8 (15.4) | |
| CTS—Trifurcation, n (%) | 21 (2.8) | 5 (12.5) | 16 (30.8) | |
| CTS—Single lung, n (%) | 2 (2.2) | 1 (2.5) | 1 (1.9) | |
| Bronchoplasty, n (%) | 10 (9.7) | 4 (9.3) | 6 (10) | 1 (OR = 0.9) |
| Bypass time (minutes), median (IQR) | 110 (81.5–138) | 127 (105–172) | 98 (69.5–121) | <0.001 |
| ACC time (hours), median (IQR) | 0 | 0 (0–4) | 0 | 0.02 |
| Operation time (hours), median (IQR) | 4.5 (4–5) | 4.5 (4–5.9) | 4 (3.5–5) | 0.01 |
| Mechanical ventilation time (hours), median (IQR) | 96 (56–157) | 138 (96–224) | 72 (48–117) | <0.001 |
| Length of ICU stay (days), median (IQR) | 8 (5–14) | 12 (9–18.5) | 6 (5–9) | 0.001 |
| Length of hospital stay (days), median (IQR) | 20 (14–26) | 24 (16.5–36) | 16 (13–21.5) | 0.004 |
| Reoperation, n (%) | 6 (5.8) | 5 (11.6) | 1 (1.7) | 0.08 (OR = 0.13) |
| Sepsis, n (%) | 11 (10.7) | 9 (20.9) | 2 (3.3) | 0.007 (OR = 0.13) |
| Early mortality, n (%) | 8 (7.8) | 6 (14) | 2 (3.3) | 0.06 (OR = 0.21) |
| Overall mortality, n (%) | 15 (14.6) | 12 (27.9) | 3 (5.0) | 0.002 (OR = 0.13) |
| Variables | Overall (N = 103) | Group A (preoperative unstable condition) (n = 43) | Group B (preoperative stable condition) (n = 60) | P-value |
|---|---|---|---|---|
| Age at operation (months), median (IQR) | 7 (3.6–13.4) | 3.9 (2.3–6.3) | 10.2 (6.5–19.2) | 0.002 |
| Weight (kg), median (IQR) | 6.7 (5.3–8.95) | 5.5 (4.65–6.5) | 8 (5.9–9.8) | 0.009 |
| Age groups, n (%) | ||||
| ≤2 months of age | 8 (7.8) | 6 (14) | 2 (3.3) | 0.07 |
| >2 months of age | 95 (92.2) | 37 (86) | 58 (96.7) | |
| None CHD | 31 | 11 | 20 | |
| Associated cardiac lesions, n (%) | 72 (69.9) | 32 (74.4) | 40 (66.7) | 0.39 |
| Risk category 2 | 8 | 1 | 7 | |
| Risk category 3 | 63 | 30 | 33 | |
| Risk category 4 | 1 | 1 | 0 | |
| Positive preoperative bacterial culture in the respiratory track, n (%) | 17 (16.5) | 14 (32.6) | 3(5.0) | <0.001 |
| Continuous stridor, n (%) | 90 (87.3) | 43 (100) | 47 (78.3) | 0.005 |
| Length of tracheal stenosis, median (IQR) | 4.5 (3.5–5) | 4 (3.5–5) | 5.0 (3.7–5.1) | 0.03 |
| Tracheal airway mucositis, n (%) | 43 (41.7) | 31 (72.1) | 12 (20) | <0.001 |
| Tracheal anatomy on msCT (92 patients) | ||||
| Normal, n (%) | 59/92 (64.1) | 32/40 (80) | 27/52 (51.9) | |
| Tracheal right upper lobe bronchus, n (%) | 10 (10.9) | 2 (5) | 8 (15.4) | |
| CTS—Trifurcation, n (%) | 21 (2.8) | 5 (12.5) | 16 (30.8) | |
| CTS—Single lung, n (%) | 2 (2.2) | 1 (2.5) | 1 (1.9) | |
| Bronchoplasty, n (%) | 10 (9.7) | 4 (9.3) | 6 (10) | 1 (OR = 0.9) |
| Bypass time (minutes), median (IQR) | 110 (81.5–138) | 127 (105–172) | 98 (69.5–121) | <0.001 |
| ACC time (hours), median (IQR) | 0 | 0 (0–4) | 0 | 0.02 |
| Operation time (hours), median (IQR) | 4.5 (4–5) | 4.5 (4–5.9) | 4 (3.5–5) | 0.01 |
| Mechanical ventilation time (hours), median (IQR) | 96 (56–157) | 138 (96–224) | 72 (48–117) | <0.001 |
| Length of ICU stay (days), median (IQR) | 8 (5–14) | 12 (9–18.5) | 6 (5–9) | 0.001 |
| Length of hospital stay (days), median (IQR) | 20 (14–26) | 24 (16.5–36) | 16 (13–21.5) | 0.004 |
| Reoperation, n (%) | 6 (5.8) | 5 (11.6) | 1 (1.7) | 0.08 (OR = 0.13) |
| Sepsis, n (%) | 11 (10.7) | 9 (20.9) | 2 (3.3) | 0.007 (OR = 0.13) |
| Early mortality, n (%) | 8 (7.8) | 6 (14) | 2 (3.3) | 0.06 (OR = 0.21) |
| Overall mortality, n (%) | 15 (14.6) | 12 (27.9) | 3 (5.0) | 0.002 (OR = 0.13) |
ACC time: aortic cross-clamp time; CHD: congenital heart disease; CTS: congenital tracheal stenosis; ICU: intensive care unit; IQR: interquartile range; msCT: multislice computed tomography.
Impact of an unstable preoperative condition on perioperative data and hospital complications
Tracheal airway mucositis was significantly more common in group A than in group B [31/43 (72.1%) vs 12/60 (20%), P < 0.001], following the higher incidence of a positive preoperative nasopharyngeal culture in group A than in group B [14/43 (32.6%) vs 3/60, (5%), P = 0.003]. The bypass time (P = 0.0003), aortic cross-clamp time (P = 0.02), operation time (P = 0.01), mechanical ventilation time (P < 0.001), length of intensive care unit stay (P = 0.001) and length of hospital stay (P = 0.004) were significantly longer in patients in group A than in group B. The incidence of postoperative sepsis was significantly higher in group A than in group B [9/43 (20.9%) vs 2/60 (3.3%), P = 0.007]. However, postoperative pneumonia was more common in group A than in group B, but there was no statistically significant difference [17/43 (39.5%) vs 15/60 (25%), P = 0.13]. Six patients needed early reoperation (5 patients with anastomotic dehiscence and 1 with residual tracheal stenosis); the reoperation rate in group A was higher than that in group B, but the difference was not significant [5/43 (11.6%) vs 1/60 (1.7%), P = 0.07].
Risk factors related to prolonged postoperative ventilation were analysed. Cox univariate analysis revealed that patients in group A (P = 0.008; HR 2.57; 95% CI 1.28–5.16%) were at significant risk for prolonged postoperative ventilation.
Effects of preoperative conditions on mortality and risk factors for overall mortality
The median follow-up period was 28.4 months (IQR 15.3–47.3 months), with the median follow-up of group A was 26.5 months (IQR 8.45–53.2 months) and in group B was 24.1 months (IQR 27.1–51.2 months). There were 3 patients who lost follow-up (2 in group A and 1 in group B). There were 8 in-hospital deaths (7.8%, 8/103), and the early mortality in group A was higher (but not significantly) than the early mortality in group B [6/43 (14%) vs 2/60 (3.3%), P = 0.06]. The causes of in-hospital mortality were respiratory distress associated with nosocomial infection in 4 patients, respiratory failure secondary to accumulating positive postoperative fluid balance combined with nosocomial infection in 3 patients and septic shock due to mediastinitis with anastomotic dehiscence in 1 patient. The linearized risk of deaths over time was 5.1 deaths per 100 patients per year, with group A reporting 11.3 deaths per 100 patients per year versus group B’s 3.6 deaths per 100 patients per year.
There were 7 late deaths (6.8%, 7/103) at a median of 3 months (IQR 1.7–5.1 months) after discharge from the hospital, resulting in overall survival rates at 1 year and 5 years of 86.4% (95% CI 80–93.3%) and 85.3% (95% CI 78.7–92.5%), respectively (Fig. 2). Four patients died of pneumonia, 2 had unknown causes of death without any symptoms before discharge and 1 died of respiratory distress due to postoperative severe tracheobronchial malacia. The overall mortality in group A (Fig. 3) was significantly higher than that in group B [27.9% (12/43) vs 5% (3/60), P = 0.001]. The risk factors for overall mortality according to Cox multivariate regression analysis were prolonged ventilation time (P = 0.024; HR 3.86; 95% CI 1.20–12.49%), additional bronchoplasty (P = 0.004; HR 5.86; 95% CI 1.72–19.31%) and tracheal airway mucositis (P = 0.011; HR 5.67; 95% CI 1.51–21.31%) (Table 2).
Overall survival of slide tracheoplasty.
Overall mortality between group A and group B.
Multivariate Cox regression analyses for overall mortality
| Variables | HR | 95% CI | P-value |
|---|---|---|---|
| Risk factor analysis for overall mortality | |||
| Prolonged mechanical ventilation (>5 days) | 3.86 | 1.20–12.48 | 0.024 |
| Concomitant bronchial plasty | 5.77 | 1.72–19.31 | 0.004 |
| Tracheal airway mucositis | 5.67 | 1.51–21.31 | 0.010 |
| Variables | HR | 95% CI | P-value |
|---|---|---|---|
| Risk factor analysis for overall mortality | |||
| Prolonged mechanical ventilation (>5 days) | 3.86 | 1.20–12.48 | 0.024 |
| Concomitant bronchial plasty | 5.77 | 1.72–19.31 | 0.004 |
| Tracheal airway mucositis | 5.67 | 1.51–21.31 | 0.010 |
CI: confidence interval; HR: hazard ratio.
Multivariate Cox regression analyses for overall mortality
| Variables | HR | 95% CI | P-value |
|---|---|---|---|
| Risk factor analysis for overall mortality | |||
| Prolonged mechanical ventilation (>5 days) | 3.86 | 1.20–12.48 | 0.024 |
| Concomitant bronchial plasty | 5.77 | 1.72–19.31 | 0.004 |
| Tracheal airway mucositis | 5.67 | 1.51–21.31 | 0.010 |
| Variables | HR | 95% CI | P-value |
|---|---|---|---|
| Risk factor analysis for overall mortality | |||
| Prolonged mechanical ventilation (>5 days) | 3.86 | 1.20–12.48 | 0.024 |
| Concomitant bronchial plasty | 5.77 | 1.72–19.31 | 0.004 |
| Tracheal airway mucositis | 5.67 | 1.51–21.31 | 0.010 |
CI: confidence interval; HR: hazard ratio.
A competing risk analysis was performed, with the events being death without reoperation, survival and freedom from reoperation, and needed reoperation for tracheal stenosis or anastomosis dehiscence. The competing risk curve showed that at 1 year after surgery, the cumulative incidence of survival and freedom from reoperation, death without reoperation and reoperation were 3.9% (95% CI 1.3–9%), 10.7% (95% CI 5.6–17.5%) and 3.9% (95% CI 1.3–9%), respectively. At 5 years after surgery, these values were 68% (95% CI 57.9–76.1%), 11.7% (95% CI 6.3–18.7%) and 5.8% (95% CI 2.4–11.5%), and at 7 years after surgery, these values were 81.6% (95% CI 72.3–88%), 11.7% (95% CI 6.3–18.7%) and 5.8% (95% CI 2.4–11.5%), respectively. The overall cumulative incidence of mortality at 1, 5 and 7 years were: 3.8% (95% CI 1.2–8.9%), 8.7% (95% CI 4.3–15.2%) and 9.7% (4.9–16.4%), respectively.
The cumulative incidence of survival with freedom from reoperation in group A at 70 months after surgery was 58.1% (95% CI 41.5–71.6%), that of death without reoperation was 20.9% (95% CI 10.2–34.2%) and that of reoperation was 11.6% (95% CI 4.2–23.3%); those of group B were 90% (95% CI 78.2–95.6%), 5% (95% CI 1.3–12.7%) and 1.7% (95% CI 0.1–7.9%), respectively (Fig. 4). The cumulative incidence of mortality in group A at 1, 5 and 7 years were 25.6% (95% CI 13.6–39.3%), 27.9% (95% CI 15.4–41.8%) and 27.9% (95% CI 15.4–41.8%), respectively versus group B with consistent of 5% (95% CI 1.3–12.7%) at 1, 5 and 7 years. Fine and Gray’s model for cumulative incidence across groups revealed that being in group A was a highly significant risk factor for overall death events (P = 0.023; SHR 4.5; 95% CI 1.23–16.4) relative to being in group B.
Cumulative incidence between groups of survival with freedom from reoperation revealed that being in group A was a highly significant risk factor for late death events.
Follow-up outcomes
Twenty-three patients underwent in-hospital bronchoscopy, with 10 having a normal trachea diameter without malacia, 6 having tracheomalacia, 3 having residual bronchus stenosis, 2 requiring emergent bronchoscopy suctioning after developing tracheal stenosis from sputum blockage and blood clots in the postoperative period, 1 having anastomotic dehiscence and 1 having a trachea–oesophagus fistula. Bronchoscopy was indicated during the follow-up period in 6 symptomatic patients, with 4 requiring reinterventions, including 2 who underwent balloon dilatation and 2 who underwent laser removal of granuloma. Four patients had continuous stridor at the latest follow-up, with 2 patients in each group.
Fine and Gray’s model revealed that patients who had a significantly higher ratio of CTS/tracheal length (>0.8) were highly significant risk factors for reintervention (P = 0.05; SHR 3.55; 95% CI 1–12.6).
DISCUSSION
Surgical treatment for CTS with complete tracheal rings has a history of more than 40 years following the 1st successful attempt in 1982 by Kimura et al. [16]. Since then, many other techniques have been developed to achieve proper surgical treatment results for this lethal congenital disease, including autologous pericardial patches [17], STP [18], tracheal autografts [19] and tracheal homografts [20]. Recent outcomes have shown great achievements with excellent results, with mortality rates ranging from 0% to 12% [4–8]. However, most of the published data originate from high-income countries with ample resources and well-established systems.
In this study, a wide spectrum of patients with complex CHD also presented with CTS. The number of patients with long segment CTS (81.6%), the median age of 7 months and especially the 41.7% needing ventilation support reflect the nature of this lethal congenital airway condition across many low-middle-income countries, with surgery as the only option to save the patients. Numerous patients are referred to our hospital with the greatest probability of fatality due to a lack of resources and knowledge for early CTS detection under the highest pressure of mechanical ventilation [median 29 cmH2O (IQR 24–33.5 cmH2O)] to release carbon dioxide and maintain vital signs. As a result, 8 patients (7.8%) experienced circulatory arrest as soon as they arrived at the operating theatre. They needed emergency cardiopulmonary resuscitation, which included emergency chest opening, cardiac massage and urgent placement of cardiopulmonary bypass.
Before 2016, many efforts to perform surgical repair for this lethal disease in our hospital failed due to mistakes in performing STP, which resulted in damage to the lateral blood supply of the trachea. Since 2016, our hospital’s natural history of CTS patients has dramatically changed. The early mortality rate in this study (7.8%) was comparable to that in previous studies [4, 8, 9, 21], which shows the versatility and applicability of STP to treat a heterogeneous population of CTS patients. However, our overall mortality rate (14.7%) is slightly higher than that reported by developed countries [6, 9, 22]. Competing risk analysis revealed that the need for preoperative respiratory support (group A) had a significantly greater impact on overall mortality than not needing such support (P = 0.023). The perioperative results from our investigation also highlight how complicated, challenging and rigorous it was to manage the patients in group A, although the tracheal length of group A was shorter than that of group B. Although the early mortality in group A was not significantly higher than that in group B (P = 0.06), our analysis demonstrated that the need for preoperative ventilation support had an adverse influence on the perioperative outcome that continued into the late mortality of the patients after surgery.
Our centre prefers to concomitantly repair tracheal stenosis with associated cardiovascular anomalies, regardless of the patient’s age or the complexity of the intracardiac procedure, to achieve a favourable postoperative resuscitation course. In contrast with Chiu and Kim [23], our study’s statistical analysis results support the notion that neither patient age (P = 0.421) nor associated cardiac anomalies (P = 0.359) impact either the mortality rate or prolonged mechanical ventilation following STP. Our data show that more than two-thirds of the cohort (72.8%) were infants <1 year old, and a substantial number (43.7%) of patients were <6 months old, while 70% of patients had concomitant cardiovascular disease. Manning et al. [6], Yokoi et al. [8] and Chiu and Kim [23] showed that younger age and smaller body weight are predictors of worse outcomes, including mortality and postoperative complications. In our study, multivariate analysis did not support these findings, indeed reflecting the difficulty of infection control in a low-middle-income country. The proportions of tracheal airway mucositis, positive preoperative nasopharyngeal viral or bacterial culture and postoperative sepsis were significant in our study, and tracheal airway mucositis appeared to be a predictor for overall mortality. This predictor not only affects prolonged mechanical ventilation but also has the potential to influence recurrent pneumonia and tracheomalacia and increase readmission to the hospital after STP and may have a greater detrimental effect on the long-term survival of CTS patients. According to the results of our statistical analyses, we believe that inhaled or nebulized ciprofloxacin may improve the treatment of the most severe forms of tracheal airway mucositis, especially in patients in low-middle-income countries.
Similar to previous studies [4, 6, 24], prolonged mechanical ventilation and bronchial stenosis were adverse predictors of patient survival after STP in our study. To our knowledge, many centres have recently proposed a ‘hands-off’ strategy when the stenosis extends to the bronchial region [4, 15]. Irrespective of the use of an aggressive strategy [4] or a ‘hands-off’ strategy when performing bronchoplasty, bronchial stenosis remains an independent risk factor for mortality. Our cumulative experience with bronchoplasty revealed that concomitant bronchial stenosis was passively identified during STP, making the surgical procedure more challenging. Therefore, proper technologies for the noninvasive preoperative identification of related bronchial stenosis (high-resolution 3D CT imaging or 3D-printed modelling) may be essential for establishing a precise surgical strategy for effectively relieving airway stenosis and enhancing STP results.
Limitations
This study has numerous limitations, including its retrospective, single-centre nature, the lack of information on preoperative CT scans and insufficient data on tracheoscopy due to the lack of a flexible endoscope to evaluate the malacia status of the tracheobronchial tract following surgery. Resource constraints also play an essential role in determining whether extracorporeal membrane oxygenation, special equipment or a particular medicine is beneficial in treating CTS patients. The sample size is small and the statistical power is limited, even though CTS is a rare congenital disease.
CONCLUSIONS
Preoperative respiratory distress produced unfavourable outcomes for patients after STP. Consequently, early detection of CTS and aggressively performed STP are important to improve the long-term survival of this disease. The intermediate-term survival for patients who undergo STP in the developing world is good and may be improved by preventing tracheal airway mucositis; however, bronchial stenosis remains a challenge to STP.
ACKNOWLEDGEMENTS
The authors would like to express their gratitude to Dr Pezzella AT, Dr Chang Haibo and Dr Jinfen Liu for their kind advisors of surgical techniques and perioperative care for STP to the VNCH team.
FUNDING
No financial support was provided for this study.
Conflict of interest: none declared.
DATA AVAILABILITY STATEMENT
All relevant data are within the manuscript and its supporting information files.
INFORMED CONSENT STATEMENT
According to our National Legislation and Institutional Requirements with retrospective study, informed consent was waived.
Author contributions
Kien Trung Nguyen: Data curation; Validation; Writing—review and editing. Anh Thi Van Nguyen: Data curation; Formal analysis; Validation; Writing—review and editing. Vinh Quang Tran: Investigation; Visualization; Writing—review and editing. Yen Thi Nguyen: Investigation; Visualization; Writing—review and editing. Chuong Thanh Le: Investigation; Visualization; Writing—review and editing. Thuc Van Dang: Investigation; Visualization; Writing—review and editing. Tae-Gook Jun: Investigation; Visualization; Writing—review and editing. Truong Ly Thinh Nguyen: Conceptualization; Investigation; Project administration; Visualization; Writing—review and editing.
Reviewer information
European Journal of Cardio-Thoracic Surgery thanks Andrea Zuin, Nagarajan Muthialu and the other anonymous reviewers for their contribution to the peer review process of this article.
Presented as an e-poster at the STS 60th Annual Meeting, San Antonio, TX, USA, 27–29 January 2024.
REFERENCES
ABBREVIATIONS
- CHD
Congenital heart diseases
- CI
Confidence interval
- CTS
Congenital tracheal stenosis
- HR
Hazard ratios
- IQR
Interquartile range
- STP
Slide tracheoplasty
