Predictors of Anastomotic Complications After Resection and Anastomosis for Tracheal Stenosis

Abstract

Resection and anastomosis is an effective option for the management of complex tracheal stenosis, however, it’s not without the complications. This study aims at evaluating various factors predicting anastomotic complications after trachea resection and anastomosis. This is a retrospective analysis of database from a dedicated thoracic surgical unit in New Delhi, India over 7 years. An analysis of demographic details, perioperative variables including complications were carried out. Analysis of various factors predicting anastomotic complications was performed. Out of 65 patients in the study, 49 (75.3%) were males and 16 (24.7%) were females. Median age of the patients was 31 years. Stenosis was cervical in 80%, cervico-thoracic in 15.4% and thoracic in 4.6% of patients. Median length of stenosis was 2.9 cm (1–4.2). 53 (81.6%) patients had some kind of preoperative intervention, where as rest 12 (18.4%) patients had no intervention at all. Out of 65 patients, 26 (40%) had crico-tracheal anastomosis while 39 (60%) had tracheo-tracheal anastomosis. Median length of resected tracheal segment was 3.3 cm (1–5). Overall complication rate (anastomotic + non-anastomotic) was 18.4% in which anastomosis related were in 4 (6.1%) patients. Resection of tracheal segment $\ge$ 3.5 cm, presence of diabetes mellitus and pre-operative use of corticosteroids were statistically significant factors for the onset of complications. Perioperative mortality rate was 1.5% (n = 1). Length of resection > 3.5 cm, presence of diabetes mellitus and pre-operative prolonged use of corticosteroids were significant predictors for the anastomotic complications.

Introduction

Tracheal stenosis (TS) is a complex condition which results from an altered inflammatory response to mucosal injury, with or without cartilage involvement. Healing by excessive circumferential scar formation leads to fibrosis and luminal narrowing, which endangers the airway and can be life threatening. Causes of benign tracheal stenosis are prolonged tracheal intubation, tracheostomy, trauma, corrosive injury, Inflammatory conditions like Wegener’s Granulomatosis, Rheumatoid Arthritis and idiopathic. Amongst these, post intubation tracheal stenosis (PITS) or post tracheostomy tracheal stenosis (PTTS) are the commonest with reported incidence of 6–21% and 0.6–21% respectively [1, 2]. TS can be classified as “simple”, which is a short length (< 1 cm) structure with only mucosal involvement without cartilage destruction or as “complex”, which is long segment lesion with cartilage destruction. Bronchoscopic interventions provide immediate relief to all, can be curative in simple stenosis and serve as an excellent bridge to surgery in complex cases. Surgical resection with clear margins and anastomosis between healthy walls remains the most effective long-term management for complex cases [3–7].

In spite of being very effective, this procedure is not without complications, which can be anastomosis related or non-anastomotic complications. These complications not only prolong the hospital stay, but can also be life threatening, especially the anastomotic ones [8, 9]. The knowledge of factors predicting such complications is of paramount importance to improve the surgical outcomes. This study aims to evaluate the factors predicting the postoperative anastomotic complications after resection and anastomosis for tracheal stenosis.

Methods

A retrospective review of prospectively maintained database of tracheal stenosis cases operated between March 2012 and December 2019 in the Thoracic surgery unit of a tertiary care hospital were included in the study. Institutional ethics committee approval was sought and individual patient consent was taken. Data was obtained from the preoperative and hospital records, operation notes, follow up visits data and telephonic conversation with the patients. TS patients with cricoid cartilage involvement, who required surgical correction of cricoid also, were excluded from the study.

Indications of Surgery

1. Symptomatic patient with complex stenosis,

2. Reduction in the tracheal diameter of more than 50% or patient having dyspnea or stridor,

3. Failure of tracheostomy decannulation due to presence of TS,

4. failure of conservative or bronchoscopic management.

Pre-operative Evaluation

Demographic data, details of present illness including duration of symptoms and history of co-morbidities were recorded. Detailed medication history, especially steroid intake history was asked for. History of previous interventions including the frequency was also noted in detail. A detailed clinical evaluation was followed by a chest X-ray, CT-scan of neck and upper chest with 3D reconstruction and bronchoscopy to evaluate the site, length and degree of stenosis, its distance from the vocal cords, the status of the vocal cords, the condition of the tracheal mucosa above and below the stenosis, any tracheomalacia and any other associated pathology. Based on the investigations, the site of stenosis was labelled as Subglottic—if the proximal end of the lesion was above lower border of cricoid cartilage, Cervicothoracic—if the lower end of the lesion was at or below the sternal notch, Cervical—if lesion was between the first two and Thoracic—if the whole or majority of the lesion was below sternal notch.

Pre-operative Optimization

No patient was taken up for surgery as an emergency. All were elective surgery, performed after adequate preparation and optimization. In patients with acute airway obstruction, bronchoscopic dilatation was undertaken to give immediate relief. However, stenting was avoided as it extends the damage to tracheal mucosa and affects the outcome [10]. If the patients were on steroids, the dose was tapered gradually and preferably stopped. Nutritional correction and correction of other co-morbidities was also done in a planned way. During pre-operative optimization, patients at risk of worsening of symptoms underwent bronchoscopic dilatation to avoid catastrophe during the waiting period.

Anesthesia

All patients were operated under general anaesthesia. Airway was secured with endotracheal tube of appropriate size passed through the stenosis, always under fibre-optic guidance. We use “paediatric uncuffed endotracheal (ET) tubes” in severe stenosis. “Telescoping” of paediatric ET tubes into another larger calibre ET tube was done to gain extra length. In case a tracheostomy was already there, we changed into a new, sterile tube at surgery and the same was used for anaesthesia by connecting a sterile circuit.

Principles of Surgery [3,11]

1. Avoid tension at the anastomotic site.

2. Prevent devascularization of tracheal ends—by restricting the length of circumferential dissection.

3. Protect Recurrent Laryngeal nerve and oesophagus.

Surgical Procedure

Operative Approach

Surgery was done in supine position with neck extension.

• In cervical tracheal stenosis—a collar incision was given, 1 inch above the supra-sternal notch to access the cervical trachea.

• In cervico-thoracic tracheal stenosis—along with collar incision, upper partial sternal split was done for better access.

• In thoracic tracheal stenosis—complete mid-sternotomy with neck incision or posterolateral thoracotomy was used. The trachea was exposed, and the stenotic segment was identified and resected.

Surgical Technique

Superior and inferior sub-platysmal flaps were raised. Strap muscles and thyroid isthmus divided in midline to reach trachea. Sharp bipolar dissection was done at stenotic segment staying very close to tracheal cartilage and stenosed tracheal segment was looped. Unnecessary lateral dissection at the healthy tracheal segments was avoided to preserve collateral vascular supply. Utmost care was taken to preserve recurrent laryngeal nerve, oesophagus.

Release Manoeuvres Performed

• In cervical tracheal stenosis—Cervical flexion, Supra and infra-hyoid release, anterior cervico-mediastinal mobilization was performed.

• Cervico-thoracic tracheal stenosis—Along with cervical flexion, hyoid release and anterior mediastinal mobilization, thoracoscopic pericardial and hilar mobilization and division of inferior pulmonary ligament were performed if necessary.

• Thoracic tracheal stenosis—Complete hilar release, pericardio-phrenic release, inferior pulmonary ligament release was performed. Cervical flexion and hyoid release were also selectively performed in long-segment stenosis.

Anterior tracheotomy was performed over the stenosed segment, and the tracheal lumen was assessed for the extent of stenosis. In patients with tracheostomy, the stoma was also resected with the stenotic segment. A flexo-metallic tube was passed into the distal trachea for cross-field ventilation. An end-to-end tracheal anastomosis was performed—the posterior and lateral walls were anastomosed with continuous 3–0 PDS suture and anterior wall was anastomosed with interrupted sutures. In case of the anastomosis being very close to innominate artery, ipsilateral sternocleidomastoid muscle was placed between them to avoid catastrophic event of tracheo-innominate fistula. After completion, the anastomosis was always checked for air leak up to pressure of 20–30 cm H2O. At the end of the procedure, all patients were extubated on the table under bronchoscopic guidance. We routinely use the Chin-to-chest guardian stitch in all cases. Vocal cord status was always checked at extubation.

Post-Operative Care

Nasogastric tube feeds and aggressive chest physiotherapy was commenced as early as possible. Patients were instructed not to hyperextend his/her neck in the postoperative period. Tube feeds were continued till the time patient accepted liquids and solids orally without signs of aspiration. None of the patient required a permanent feeding tube.

Postoperative complications were classified as anastomotic related or non-anastomotic. Anastomotic complications included anastomotic dehiscence, restenosis or granulation tissue at the anastomotic line. All other complications were considered non-anastomotic.

Follow-Up

Patients were asked to follow up after 1 week, 1 month, 6 months and then yearly or SOS. Check bronchoscopy was done at 6 months to assess the tracheal lumen and condition of anastomosis.

Statistical Analysis

Statistical testing was conducted with the statistical package for the social science system version SPSS 23.0. Continuous variables were presented as mean ± SD or median (IQR). Categorical variables were expressed as frequencies and percentages. The comparison of normally distributed continuous variables between the groups was performed using Student’s t test. Nominal categorical data between the groups were compared using Chi-squared test or Fisher’s exact test as appropriate. Non-normal distribution continuous variables were compared using Mann Whitney U test. Multivariate analysis was done using linear regression model. For all statistical tests, a p value less than 0.05 was taken to indicate a significant difference.

Results

Demographic Details and Pre-Operative Variables

Out of 65 patients in the study, 49 (75.3%) were males and 16 (24.7%) were females. Majority of the patients were below 40 years of age (n = 51) with median age of 31 years. Trauma was the major cause of intubation (47.7%) followed by pneumonia (26.1%) with a median intubation duration of 11 days (range of 2–36 days). Two patients had no history of intubation as one of them had stenosis related to airway tuberculosis and other related to Rheumatoid arthritis. Majority (84.6%) of patients had no co-morbidities. Four patients were using corticosteroids for long duration pre-operatively.

The site of stenosis was cervical in location in majority of patients (80%), where as 15.4% of patients had cervico-thoracic stenosis and 4.6% had thoracic tracheal stenosis. The severity of the stenosis was grade III in majority of patients (66.1%), where as 26.1% had grade II stenosis and 7.6% had grade IV stenosis as per Myer-cotton grading. As per the CT scan measurements, the median length of stenosis was 2.9 cm with a range of 1–4.2 cm. None of our patients had associated tracheomalacia. Fifty-three (81.6%) patients had some kind of preoperative intervention, where as rest 12 (18.4%) patients had no intervention at all. Bronchoscopic dilatation was the commonest procedure (56.9%) followed by tracheostomy (29.2%). Out of the total 37 patients who had bronchoscopic dilatations, 16 patients underwent multiple procedures over a period of time. So, in total 37 (63.8%) patients underwent 86 dilatation procedures with an average of 2.3 procedures per patient (Table 1).

Table 1.

Demographic details and pre-operative variables (n = 65)

Characteristics	Number of patients (Percentage)
Male: Female	49:16 (75.3: 24.7)
Median age in years (Range)	31 (10–81)
Co-morbidities
Diabetes (DM)	4 (6.1%)
Hypertension (HTN)	3 (4.6%)
Hypothyroidism	3 (4.6%)
None	55 (84.6%)
Cause for initial Intubation
Trauma	31 (47.7%)
Pneumonia/ARDS	17 (26.1%)
Others	15 (23%)
No history of Intubation	02 (3.1%)
Median Duration of intubation in days (Range)	11 (2–36)
Site of stenosis
Cervical trachea	52 (80%)
Cervico-thoracic trachea	10 (15.4%)
Thoracic trachea	3 (4.6%)
Grade of stenosis (Myer–Cotton)
I	Nil
II	17 (26.1%)
III	43 (66.1%)
IV	5 (7.6%)
Median length of segment affected (in cm)	2.6 (1–4.2)
Pre-operative intervention
Yes	53 (81.6%)
No	12 (18.4%)
Type of pre-operative intervention
Bronchoscopic measures and dilatation	22
Silicone stenting	8
Tracheostomy	19
T-tube insertion	4

Peri-Operative Variables

Induction of anaesthesia was through tracheostomy in 19 (29.3%) and rest 46 (70.7%) through orotracheal tube placed under bronchoscopic guidance. Out of total 65 tracheal resections, 26 (40%) patients had crico-tracheal anastomosis while 39 (60%) had tracheo-tracheal anastomosis. In patients, who had tracheostomy, the segment having tracheostoma was also resected in toto. Median length of resected tracheal segment was 3.3 cm in a range of 1–5 cm. At the end, the anastomosis was covered with local vascular tissue (thyroid gland and strap muscles). However, pedicled pectoralis major muscle flaps were used in 9 (14%) patients when the anastomosis was close to vascular structures. All patients were extubated at the end of the surgery.

Overall complication rate (anastomotic + non-anastomotic) was 18.4%. Anastomotic complications developed in 4 (6.1%) patients. Two patients (3%) had anastomotic dehiscence, out of which 1 patient was re-explored in OT, and tracheostomy was inserted through the anterior dehiscence, which was removed later successfully without further complications. Another patient was managed conservatively with opening of the wound and regular dressing. One (1.7%) patient who developed granulation at site of anastomosis at 4 months was treated successfully with bronchoscopic management. Only death (1.5%) in the series happened in a patient due to tracheo-vascular fistula leading to haemorrhage and asphyxia.

Eight patients (12.3%) developed non anastomotic complications. Wound infection being the commonest (n = 4. 6.1%), which was managed with regular dressings. Vocal cord oedema or paralysis are important with respect to functional outcome. One (1.5%) patient developed vocal cord oedema while 1 (1.5%) developed reversible vocal cord paralysis. All of them recovered completely. One (1.5%) patient developed temporary quadriparesis due to excessive neck flexion, which was managed conservatively with posture correction and intravenous methyl prednisolone with immediate and complete recovery. Another (1.5%) patient developed pneumothorax on right side on day 1, for which a chest tube had to be inserted (Table 2).

Table 2.

Peri-operative variables (n = 65)

Characteristics	Number of patients (Percentage)
Route of induction of anaesthesia
Orotracheal intubation	46 (70.7%)
Tracheostomy	19 (29.3%)
Approach
Transverse cervical (TC) incision alone	52 (80%)
TC + Partial sternotomy	7 (10.7%)
TC + Partial sternotomy + Right thoracotomy/thoracoscopy	3 (4.6%)
Mid sternotomy	2 (3%)
Right postero lateral thoracotomy	1 (1.5%)
Median length of trachea resected (cms)	3.3 (1–5)
Type of anastomosis
Crico-tracheal	26 (40%)
Tracheo-tracheal	39 (60%)
Total number of complications	12 (18.4%)
Anastomotic complications	4 (6.1%)
Dehiscence	2 (3%)
Granulation tissue	1 (1.5%)
Dehiscence with Tracheo-vascular fistula	1 (1.5%)
Non-anastomotic complications	8 (12.3%)
Wound infection	4 (6.1%)
Temporary Quadriparesis	1 (1.5%)
Pneumothorax	1 (1.5%)
Vocal Cord paralysis	1 (1.5%)
VC edema	1 (1.5%)
Peri-operative death(s)	1 (1.5%)
Median hospital stay in days (Range)	7 (2–30)

Factors Affecting Anastomotic Complications

On analysis of factors for development of anastomotic complications, resection of tracheal segment $\ge$ 3.5 cm, presence of diabetes mellitus and pre-operative use of corticosteroids had significantly more chances of developing anastomotic complications (Table 3).

Table 3.

Factors affecting anastomotic complications

Variables	Characteristics	Anastomotic complications		P value
		Yes	No
Length of resected segment	$\ge$ 3.5 cm	4	26	0.04
	< 3.5 cm	0	35
Pre-op intervention (Dilatation/stenting/Tracheostomy)	Yes	4	49	1.0
	No	0	12
Tracheostomy status at surgery	Yes	2	17	0.57
	No	2	44
Crico-tracheal anastomosis	Yes	3	23	0.29
	No	1	38
Diabetes mellitus	Yes	2	2	0.01
	No	2	59
Pre-operative steroid usage	Yes	2	2	0.01
	No	2	59

Discussion

The aetiology and pathogenesis of tracheal stenosis is well established. Despite this the cases of Post-Intubation or Post-Tracheostomy stenosis continue to rise because of increasing intensive care facilities and a greater number of people requiring artificial ventilation due to various reasons [12–14]. A thorough evaluation with CT neck and chest with 3D airway reconstruction and bronchoscopy is a must in all patients. CT is supplementary to bronchoscopy as it provides exact length of stenosis when bronchoscope cannot be negotiated. It also provides additional information about wall of the airway, distal airway, lung parenchyma or any other co-existing pathology in the neck or chest [15].

The management of TS can be broadly categorised into bronchoscopic intervention and surgical intervention. Bronchoscopic interventions are effective in case of simple stenosis, however they lead to high rate of recurrence in complex stenosis, in which case surgery is the definitive treatment option [16]. In complex stenosis, the role of bronchoscopic interventions should be considered only in case of 1.an unfit patient, 2. acute airway obstruction to tide over the crisis or in patients who needs pre-operative optimization, 3. management of anastomotic complications. Stenting should be avoided in a potentially operable patient as it damages mucosa, causes local inflammation leading to granulations, increases the extent of mucosal injury and thus length of resection [17]. So, patients with complex stenosis (not in acute airway obstruction) should be offered surgery per primum. Surgery should be avoided at all costs in an emergency situation and in such cases either bronchoscopic interventions if available or a tracheostomy needs to be done, the tube should be inserted through the stenotic segment to reduce the length of resection required later [11]. The aim of the surgery is to provide a healthy airway of adequate diameter so that the patient returns to his previous activity level with normal vocalization and normal swallowing. The surgical principles enunciated by Grillo et al., Pearson et al., and Couraud et al.are extremely important and should be strictly adhered to [3, 11, 18].

The overall morbidity after tracheal resection varies from 17–45% and mortality ranges from 0–2.4% [19, 20]. In our series also, the overall complication rate was 18.4% and perioperative mortality was 1.5%, which was similar to the reported literature. Although tracheal resection and anastomosis is an effective procedure, anastomotic complications are not uncommon and are multifactorial [21, 22] with a reported incidence of 3.6–7.1% [3, 23]. In present study, 4 (6.1%) patients developed anastomotic complications.

Out of the many factors affecting anastomosis, only few can be modified but identifying them is of utmost important as this improves procedural planning, helps in patient counselling [17] and also offers an opportunity to modify these for better outcomes. Wright et al. performed a multivariate analysis of tracheal resection for various pathologies. They concluded that Diabetes, re-operation, longer resections (> 4 cm), need for tracheostomy before surgery, and laryngotracheal resections were associated with complications. The need for release procedure was not considered as a risk factor according to them [6] In the present series, length of resection > 3.5 cm, presence of diabetes mellitus and prolonged use of corticosteroids in the pre-operative period were risk factors for anastomotic complications.

Couraud et al. stated that half of the adult human trachea can be safely resected [11] As per Mathisen this can be done only under ideal conditions—viz a healthy patient, long slender neck, incorporation of all the manoeuvres to gain tracheal length viz neck flexion, laryngeal release, pericardial release with division and reimplantation of airways [24]. In our study resection length ≥ 3.5 cm had higher risk of developing anastomotic complication contrary to limit of 4 cm in western literature, this difference may be attributed to difference in the mean length of trachea in Indian population (Males 10.42 ± 0.5 cm and Females 9.8 ± 0.2 cm) compared to the American population (Mean 11.8 cm) [25].

Diabetes increases the incidence of anastomosis related complications by an odds ratio of 2.7. Diabetes causes impairment of microvascular circulation in the mucosa causing decreased anastomotic perfusion [26] Prolonged use of steroids also affects tissue healing and has potential to cause anastomotic dehiscence [17]. So, tapering of steroids should be considered as a critical concern in such patients.

Presence of tracheostomy or t-tube at the time of surgery was also reported as an independent risk factor for anastomotic complications [6]. Presence of such appliance affects the anastomotic process in several ways. 1. Presence of tracheostomy/T-tube encourages the colonisation of microorganisms—increased inflammation, 2. The site of tracheostomy may become malacic segment, 3. Peri-tracheostomy adhesions limits the mobility of the trachea during surgery and causes tension on the anastomosis, 4. need for inclusion of tracheostoma in the resection margins, increases the length of resectable tracheal segment. However, in our series presence of tracheostomy/ T-tube didn’t significantly affect the anastomotic complications. This can be explained by the small sample size and performance of tracheostomy through the stenotic segment in most of the cases.

Limitations—It is retrospective in nature and is based on a single institution’s experience with limited number of patients as compared to other studies. Mild heterogenicity of the study population also affects the study outcome. A larger data set would be beneficial to draw definitive conclusions.

In conclusion, tracheal resection and anastomosis is the treatment of choice for complex benign tracheal stenosis. This procedure is safe and effective; however, incidence of anastomotic complications is not uncommon. Length of resection > 3.5 cm, presence of diabetes mellitus and prolonged use of corticosteroids in the pre-operative period were risk factors for anastomotic complications. Hence, all modifiable risk factors should be identified and optimized pre-operatively, such as optimization of co-morbidities i.e., diabetes mellitus, and tapering of cortico-steroids. Appropriate patient selection, pre-operative optimization, rigid adherence to the principles of tracheal surgery, vigilant postoperative care and multidisciplinary coordination are the key for best postoperative outcomes.