Management of benign airway stenosis—predictors of tracheal resection

Introduction

Benign airway stenoses are complex disorders that present with progressive dyspnoea, stridor, and significant respiratory distress (1). In adults, the most common causes are intubation/post-tracheostomy: incidence 1:200,000 adults annually, more even sex distribution and higher rates of co-morbidities, idiopathic: incidence 1:400,000, mostly female, Caucasian, otherwise healthy, systemic disease, trauma, radiotherapy, and infection (2). They are often misdiagnosed as asthma or chronic obstructive pulmonary disease (COPD) and delays in diagnosis and treatment of up to 4 years have been reported (1,3). The management is complex, often requiring a multidisciplinary approach with a close collaboration between thoracic surgeons, ear, nose, and throat (ENT) specialists, various medical specialities, and anaesthetists (2). These conditions have a high recurrence rate and despite the plethora of endoscopic and open techniques available for treatment, the outcomes are variable (3). Grillo was the first to describe the surgical management of this unique and complex cohort of patients (2), ranging from simple tracheal resection (TR) to complex cricotracheal and laryngotracheal reconstruction. This makes shared decision making difficult, with patients’ desire to have good quality of life (QOL) without fear of recurrence of symptoms, and the surgeon to consider the complexity of the definitive operative patient and its timing. Large scale national data has demonstrated the feasibility of integrating a network of surgeons to facilitate treatment comparisons and help inform treatment-related decisions (4).

In this manuscript, we present our series of benign airway stenoses that have been managed both endoscopically and surgically. We sought to determine whether there were any strong predictors of the need for surgical resection based on our experience. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-727/rc) (5,6).

Methods

This retrospective observational analysis was conducted at a single centre, serving a regional, national, and international population and treating almost 2.2 million patients per year. Between Jan 2008 and June 2022, data were gathered concerning all patients with benign tracheal stenoses who have undergone intervention at Queen Elizabeth Hospital Birmingham (QEHB). We have a regional airway interest group meeting (comprised of respiratory physicians with an airway interest, ENT surgeons, thoracic surgeons, radiologists and physiotherapists) which meets monthly where patients with complex airway pathology are discussed. All patients are seen regularly for evaluation of symptomatology and in those who are unable to function properly as a result of the stenosis, undergo a screening airway assessment. The main criterion for initial airway assessment is the presence of symptoms.

The time period was dictated by the availability and completeness of data from the electronic patients’ records (EPR) and the national audit database. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was registered with the local NHS Trust (internal audit code 14333), and no formal ethical approval was required given that there was no direct patient contact in this study. Patient involvement and therefore consent was not a part of this study.

Our primary outcome measure was to ascertain the need for TR as a time-dependent outcome. Time to surgery was started from the point of initial referral to a specialist regardless of the institution. We retrospectively ascertained clinical and demographic data at admission and during their inpatient stay in this time period. The data included demographics, past medication history, smoking status, and co-morbidities. We reported the characteristics of the stenotic lesions in patients, including location, aetiology, proximity to the vocal cords, symptoms, Freitag classification (7), European Laryngological Society (ELS) score and Myer-Cotton grading. The number of pre-operative dilatations undertaken was reported and the number of patients who underwent formal TR of stenosed segment. Post-operative mortality, length of stay, symptoms at last consultation, rates of surveillance bronchoscopy and restenosis rates and their management were reported. QOL measures were assessed using the EQ5D scoring system. All outcome measures were ascertained at longest-term follow-up, this was at least 5 years in both surgical and non-surgical groups, with patient-led follow-up conducted in the non-surgical conservative group.

In addition to interrogating the entire dataset, we undertook a selective analysis of those patients with idiopathic subglottic stenosis (iSGS).

Statistical analysis

For continuous variables, results are expressed as means and standard deviations or medians and interquartile range (IQR) depending on the distribution of the data. For categorical variables, we reported counts and percentages. For continuous data, group comparison was carried out using a Student’s t-test or Mann-Whitney test depending on the distribution of data. Normality was assessed using Shapiro-Wilk testing. Group differences for categorical data were assessed using the chi squared test of independence, or Fisher’s exact test for low frequencies. Univariate analyses of risk factors associated with TR were assessed by performing individual unadjusted logistic regression analysis with inclusion of one covariate per model. A backward elimination, stepwise multivariable logistic regression analysis was performed to identify the independent predictors of TR within this dataset as a whole and sub-stratified according to those with idiopathic aetiology and pure subglottic stenosis (SGS). Time-to-event data was analysed using Cox proportional hazards modelling. Receiver operating characteristic (ROC) analysis was carried out using the pROC package and the optimal cutpoints (sensitivity/specificity balance) were determined using the cutpointr package. Survival analyses were carried out using survminer and displayed using the ggsurv/ggplot2 packages (8-11). Data were displayed using the gtsummary package. The tests were considered significant at P<0.05. Missing data were excluded from analyses, however where possible we used multiple imputation for data missing at random (12). All analyses were performed using R programming software (v4.0.3) in R studio.

Results

Study cohort

We analysed 97 patients with benign airway stenoses over a 14-year period [2008–2022]. Full demographic data are shown in Table 1. The median age of the cohort was 44 (IQR, 34–54) years with a significant female preponderance (n=79, 81%). The majority of patients (n=78, 80%) were life-long never smokers. Asthma and COPD were the most common airway pathologies in our cohort, 8.2% and 5.2%, respectively.

Disease features

SGS was the most common site of disease in the trachea (79%) (Table 2). The majority of all stenoses were idiopathic in nature (52%). Other causes included prolonged intubation (36%) and autoimmune disease (8.2%). Structural type 4 (7) was the most common type encountered (82%). We employed the Myer-Cotton grading system to quantify the degree of luminal obstruction, grade 3 disease was the most commonly encountered (87%). Stridor and shortness of breath were the common presenting symptoms. Vocal cord movement was normal in 93% of cases. At the time of referral, the median length of stenosis was 10 (IQR, 10–20) mm. Fifty-two percent of patients presented with a stenotic diameter of <5 mm, 5 patients (5.2%) presented with a 2 mm diameter, and 21 patients (22%) presented with a 3 mm diameter. Thirty-one patients (32%) had a history of previous tracheostomy. The median number of dilatations in our cohort was 3 (IQR, 2–4). Forty-one patients (42%) underwent formal laryngotracheal resection/TR.

Surgical and interventional techniques

In our cohort, we employed endoscopic and surgical techniques to deal with the benign airway disorders presenting to our unit. The same group of surgeons performed both techniques using a standardised method which has been preserved and taught over time in our unit. Pre-operative assessment of patients includes full history, examination, lung function with flow-volume loops and a detailed computed tomography (CT) scan with three-dimensional (3D) reconstruction of the airways to ensure no pathologies are missed. All patients undergo a routine airway assessment with balloon dilatation to ascertain the responsiveness of the lesion and subsequent impact on symptoms and QOL. In situations of excessive granulation tissue where patients may benefit from endoscopic resection, a diode-pumped Nd-YAG laser is employed to circumferentially resect areas of stenosis with adjuvant inhaled corticosteroids and oral proton pump inhibitors. In cases of long-segment stenoses (>6 cm/12 rings) or features of expiratory central airway collapse [tracheomalacia or expiratory dynamic airway compression (EDAC)], stents are employed, usually silicone such as the Dumon or Leufen systems.

Of the 41 patients who underwent surgery, one required hyoid release, sternal division whether full or partial was not required in any of the patients. Given that 95% of stenoses were subglottic and upper tracheal with an average length of 15 mm, it is not surprising that extensive release was not required in more cases. All patients undergo an initial screening bronchoscopy with dilatation and then review (Figure 1A). Surgical resection is undertaken in those patients who elect for surgery or those who fail to respond to dilatation (either no symptomatic relief or short interval between dilatations; either decreasing intervals or <3 months between dilatations). In patients who elect for surgery, they must all undergo an airway assessment ± dilatation in order to evaluate the aetiology and nature of the stenosis.

Figure 1 Surgical procedural steps taken during CTR. (A) Bronchoscopic findings pre-operatively. (B) Isolation and exposure of the trachea. (C) Establishing cross-field ventilation. (D) Posterior wall anastomosis. (E) Anterior wall anastomosis. CTR, cricotracheal resection.

All patients undergo a transcervical approach with division of the strap muscles and mobilisation of the trachea (Figure 1B). The airway is opened at the level of the obstruction, cross-field ventilation established, and stenotic segment excised (Figure 1C). The posterior membranous wall is reconstructed with continuous 4-0 polydioxanone suture (PDS) and anchored at both corners with an interrupted 3-0 PDS stitch (Figure 1D). The anterior wall is completed with interrupted sutures, again 4-0 PDS (Figure 1E).

Surgical sub-group

At the time of surgery, one-third of patients had a stenotic diameter less than 5 mm. The maximum length of stenosis was 30 mm (seen in 7.3% of patients), with a median length of resection of 25 (IQR, 20–30) mm. Fourteen patients (34%) had undergone pre-referral interventions. Most patients (71%) had two or more tracheal rings resected. Fifty-four percent (n=22) of patients underwent laryngotracheal resection, with resection of the anterior two-thirds of cricoid cartilage in 14 patients, and incision of the cricothyroid membrane and thyroid cartilage split in a further eight patients. All patients in this series were extubated within 24 hours of the operation and 95% were extubated in the anaesthetic room prior to transfer to the intensive care unit (ICU).

Histological findings included non-specific inflammation (n=12, 29%), chronic inflammation (n=10, 24%), scarring (n=8, 20%), fibrosis (n=7, 17%), ulceration (n=3, 7.3%) and amyloidosis (n=1, 2.4%). Median length of post-operative stay was 6 days in this cohort. Four patients experienced re-stenosis (9.8%) post-resection which required balloon dilatation. One patient underwent redo surgery. All patients had a post-operative Grillo chin stitch. One patient experienced wound infection post-operatively, and three patients experienced post-operative dysphonia (change in voice pitch and tone) and dyspnoea at long-term follow-up (median 5 years). Late death, defined as death beyond 90 days out of hospital, was seen in three patients in this subgroup, this was secondary to myocardial infarction (n=2) and urosepsis (n=1).

Conservative sub-group

Fifty-six patients (58%) with tracheal stenosis were managed conservatively. Median age was 44 (IQR, 34–52) years with a significant female preponderance (86%). The majority were never smokers (88%). A significantly higher proportion of patients in this sub-group were pure SGS compared to those who underwent TR (93% vs. 61%, P<0.001), with 63% of these being idiopathic in origin. There was no difference in vocal cord movement between the two groups. The median number of balloon dilatations carried out in this sub-group was 3 (IQR, 1–9). Prior to referral to QEHB, 71% (n=40) of this sub-group had undergone no pre-referral treatment. Balloon dilatation, laser ablation, stenting and steroid injection had been carried out in 8.9% (n=5), 16% (n=9), 1.8% (n=1), and 1.8% (n=1) respectively. and all are still under surveillance with the regional airway management group for recurrence of symptoms and need for further intervention. We determined that the median time interval between dilatations and recurrence of symptoms was 155 (IQR, 46–280) days. This tended to be shorter for autoimmune, vasculitis and idiopathic pathologies. There was no significant difference in the number of pre-referral interventions between the conservatively managed and surgically managed cohorts (P=0.40). No patients in the conservative group have required emergency surgery but instead have been expedited to surgery if they had rapid symptom recurrence.

Univariate analysis

There were no significant differences seen in demographics or co-morbidities between the surgical and non-surgical cohort of patients. A significantly higher proportion of patients who developed tracheal stenosis as a result of prolonged intubation underwent TR [71% vs. 29% (conservative group), P<0.001] (Table 3). More patients in the TR group had a higher Myer-Cotton grading; grade 3, P=0.01. Pre-operative tracheostomy rate was significantly higher in those who underwent TR (51% vs. 18%, P<0.001) (Table 3). The number of dilatations was equivocal between the surgical and non-surgical groups (P=0.30).

QOL measures [EuroQol five-dimensions (EQ5D)] were ascertained at 12 months post-intervention in the surgical arm (0.77±0.22). In those treated conservatively, this was measured at 12 months post-initial intervention (dilatation or otherwise) (0.63±0.24).

Independent predictors of TR

Utilising a multivariable backward elimination stepwise Cox proportional hazards model, we found that pre-operative tracheostomy, advanced Myer-Cotton grade, and an increased number of involved airway subsites (more complex stenoses) were the significant independent factors predictive of the need for TR (Table 4).

Using a multivariable ROC model, incorporating the four features from Table 4 (including smoking status), these risk factors collectively incurred an area under the curve (AUC) of 0.861 indicating high predictivity of surgical resection.

We allocated each of the four factors above equal weighting to generate a risk model [0–4]. This was done to facilitate ease of use in the clinical setting. Higher scores conferred an almost 5-fold higher hazard of resection [hazard ratio (HR), 4.81; 95% confidence interval (CI): 2.54–10.3; P<0.0001]. Smoking status was awarded 1 point for ever smokers (current or former) and 0 points for never smokers as per modelling previously described (13). Time to resection by risk strata (high 2–4) vs. low risk [0–1] is demonstrated in Figure 2.

Figure 2 Kaplan-Meier time to TR analysis according to different risk strata. We devised a four strata risk model to predict hazards of resection. This was significantly associated with higher hazard of resection (HR, 4.81; P<0.0001). This comprised pre-operative tracheostomy [1], higher grade stenoses [1], smoking status [1], and number of involved airway subsites [1]. Significantly shorter time to resection with increasing risk score by Kaplan-Meier log-rank analysis. TR, tracheal resection; HR, hazard ratio.

iSGS

In a subgroup of patients with iSGS (n=50), we further explored the impact of various factors on outcome, namely TR and we did not identify any further risk factors that were predictive of resection including the number of pre-operative dilatations [odds ratio (OR), 1.14; 95% CI: 0.93–1.41; P=0.20]. We further explored this relationship using a time-dependent ROC analysis to determine if there was a threshold of number of dilatations, beyond which the odds of surgery increased. ROC analysis showed that the number of pre-operative dilatations was not overtly predictive in this subgroup, AUC 0.57 with an optimal cut-off of 6 dilatations. We thence performed a “time to resection” analysis stratifying those with a high number of pre-operative dilatations (>6) vs. those with a low number (≤6) of pre-operative dilatations (Figure 3). Although this relationship was not significant in the iSGS cohort (P=0.53 by log-rank test), there is a significant drop-off in the “high dilatation” group beyond 750 days.

Figure 3 Kaplan-Meier time to TR analysis between high and low dilatation burden groups. TR, tracheal resection.

The number of affected airway subsites (2b vs. 1a) in this subgroup associated with a significantly increased hazard of resection (HR, 6.35; 95% CI: 2.24–18.0; P<0.001).

Discussion

In our series, circa less than 50% of patients were managed with resection and reconstruction. Higher Myer-Cotton grade and pre-operative tracheostomy conferred a 3- and 2-fold higher hazard of TR in our patient cohort respectively. Diligent follow-up in patients who have required prolonged mechanical ventilation and tracheostomy-facilitated weaning has been suggested as mandatory to prevent the onset of severe tracheal stenosis and need for resection (14,15). Motus et al. have suggested that endoscopic treatment undertaken in a prompt manner in these patients reduces the rate of stenosis (14) and in particular the onset of true circumferential stenosis which invariably mandates resection (16). Our experience has shown that post-tracheostomy patients are notoriously resistant to long-term relief from endoscopic balloon dilatation alone and ultimately require resection and reconstruction. The anterior tracheal border becomes densely fibrotic and scarred and as such is not responsive to dilatation. The burden of these patients is likely to increase given the large volume of patients requiring weaning from mechanical ventilation during the coronavirus disease 2019 (COVID-19) pandemic. Those patients operated on during the pandemic did not have worse peri-operative outcomes compared to those operated on outside the pandemic era. This was reported similarly by a group from Italy (17). Post-intubation tracheal stenosis patients respond well to resection and reconstruction but since the advent of high-volume, low-pressure cuffs on endotracheal tubes, the incidence of this aetiology has declined. Pre-operative tracheostomy and laryngotracheal resection in this group leads to poorer outcomes (18,19). The literature commonly reports smoking status as a risk factor for post-operative pulmonary complications following lung resection (20) however not as a predictor for resection in tracheal stenosis. Large series data have not associated smoking status as a risk factor for anastomotic failure post-TR (18,19,21,22). The majority of our cohort were never smokers and although smoking status was not a significant finding on the multivariable analysis, we chose to include this in our risk model as there is extensive evidence that cigarette smoke exposure induces airway remodelling, obstruction and impairs epithelial cell-cell integrity in the airway mucosa and this may be relevant in the progression of stenoses in the upper airways (23,24).

These patients were at the highest risk of surgery given the poor response to repeated endoscopic therapy. What we have discerned is that in those who fail initial endoscopic balloon dilatation, as reflected by early recurrence of symptoms or short interval between treatments, should be risk stratified with thorough work-up, evaluation of aetiology and complexity of the stenosis. Devising a form of risk calculator or model would be of great importance in this setting and in order to do this robustly, deriving outcomes in a larger prospective cohort is key. The risk model we have devised is intended to guide management in the clinical setting and identify patients who would benefit from early surgical intervention as opposed to repeated endoscopic therapies.

The majority of our cohort were managed conservatively with surveillance and balloon dilatation, which is described as an effective tool to improve symptoms and QOL in patients with complex stenoses (25). Four patients were stented, and our experience has shown that these patients have more durable tracheal patency compared with balloon dilatation alone however at the risk of an increased rate of perioperative complications, which has been previously reported (26).

As our data reflects, some patients were even having up to 17 balloon dilatations. One of the key questions we aimed to answer with this study was to whether there is an inflection point in terms of number of balloon dilatations beyond which TR and reconstruction is imminent given a steadily decreasing interval between serial dilatations. Time-dependent ROC analysis demonstrated this to be beyond six dilatations, however our time to surgery analysis (Figure 1) showed no significance between the two groups. The cutoff point may well be influenced by the fact that balloon dilatation was employed heavily as opposed to higher use of stents. Ultimately this was difficult to determine given that patient choice plays a huge role in the decision making for surgery. Most patients are incredibly reticent to undergo a major surgery given the potential risk for vocal cord palsy and permanent tracheostomy, and as a result opt for serial balloon dilatations up until a point where there is no symptomatic relief for a significant period of time with this intervention. This is likely to be a significant contributing factor to the lack of statistical significance with regard to number of pre-operative balloon dilatation as a predictive factor for surgery in the overall cohort and particularly the iSGS cohort. Furthermore, there is an inherent paradox in dilatations and the association with TR, dilatation or any other endoscopic intervention results in inflammation, progressive scarring, and fibrosis in the airway. This is not only circumferential, resulting in narrowing, but extends longitudinally, towards the cords thus making eventual surgical resection either complex or impossible.

In the iSGS cohort, odds of TR were higher with an increasing number of pre-operative dilatations, albeit not significant (OR, 1.14; 95% CI: 0.93–1.41; P=0.20). Large cohort data (4) has demonstrated, through weighted, propensity score-matched, Cox proportional hazards regression models that endoscopic dilatation alone was inferior to endoscopic resection with medical treatment (ERMT) (HR, 3.16; 95% CI: 1.8–5.5). Among successfully treated patients without recurrence, those treated with cricotracheal resection (CTR) had the best QOL outcomes but the worst voice outcomes at 1 year. These voice outcomes were described as a change in pitch which was only experienced by three patients in the surgical sub-group but given that all the patients were female, this was a significant cause of distress for them. This is a huge factor influencing individual decision making as previously stated and must be offset against the significantly higher recurrence rate seen with endoscopic balloon dilatation only (27).

We encountered eight patients with underlying autoimmune disease, granulomatosis with polyangiitis (n=5) and relapsing polychondritis (n=3). All of these were managed conservatively, with high frequency of serial dilatations with a median interval of 4 months between dilatations. The disease characteristics were subglottic in nature with circumferential narrowing; asymmetric stenosis was seen in two of the relapsing polychondritis patients. Disease features in this rare subgroup have been described by Catano et al. (28).

We have demonstrated good long-term outcomes and QOL after surgery however further data are needed to determine functionality. This can be achieved but vigilance in the follow-up process particularly for voice-related changes, swallowing changes and voice perception are paramount (29). Careful assessment and multi-disciplinary involvement are important, but beyond this, patient-driven decision-making is important to ensuring safe and durable long-term outcomes.

Study strengths and limitations

This study has a number of limitations; firstly, the study is retrospective making it potentially prone to interpretation and certainly a high degree of selection bias and the data was not recorded in a systematic, planned manner. A prospective study could have gathered data on pre- and post-operative functional outcomes such as voice, breathing and QOL, thus providing more information on patient-reported outcomes to aid with decision making. Furthermore, as experience in surgical resection grew, the number of surgical resections increased which may have introduced a degree of temporal selection bias. As well as patient choice, “surgeon choice” may have influenced the treatment strategy given that CTR associates with a high degree of risk and complications.

Given that we are dealing with a rare group of patients where management is not uniform between different centres, it was difficult to perform in-depth subgroup analyses. Despite these limitations, this study does demonstrate the characteristics and outcome data for this unique group of patients in a single centre and highlights the importance of managing this condition given the allied morbidity and mortality associated with it, especially when compared to other benign conditions within the thorax.

Conclusions

The challenge with this cohort of patients overall is in diagnosing the underlying condition which is heterogeneous but will influence the treatment paradigm. We found that more complex lesions, patients with pre-operative tracheostomy and previous smoking history conferred a higher hazard for resection.

Acknowledgments

Funding: None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-727/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-727/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-727/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-727/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was registered with the local NHS Trust (internal audit code 14333), and no formal ethical approval was required given that there was no direct patient contact in this study. Patient involvement and therefore consent was not a part of this study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Agrawal A, Baird BJ, Madariaga MLL, et al. Multi-disciplinary management of patients with benign airway strictures: A review. Respir Med 2021;187:106582. [Crossref] [PubMed]
2. Grillo HC. Management of idiopathic tracheal stenosis. Chest Surg Clin N Am 1996;6:811-8.
3. Ezzaouia A, Maazaoui S, Touil A, et al. Management of benign tracheal stenosis: the endoscopic approach. Eur Respir J 2022;60:3983.
4. Gelbard A, Anderson C, Berry LD, et al. Comparative Treatment Outcomes for Patients With Idiopathic Subglottic Stenosis. JAMA Otolaryngol Head Neck Surg 2020;146:20-9. [Crossref] [PubMed]
5. Vandenbroucke JP, von Elm E, Altman DG, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. Int J Surg 2014;12:1500-24. [Crossref] [PubMed]
6. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol 2008;61:344-9. [Crossref] [PubMed]
7. Freitag L, Ernst A, Unger M, et al. A proposed classification system of central airway stenosis. Eur Respir J 2007;30:7-12. [Crossref] [PubMed]
8. STHDA. survminer R package: Survival Data Analysis and Visualization. [Cited 2019 Aug 20]. Available online: http://www.sthda.com/english/wiki/survminer-r-package-survival-data-analysis-and-visualization
9. GitHub. ggsurvplot(): Plotting multiple surv objects on the same graph #195. kassambara/survminer. [Cited 2020 May 11]. Available online: https://github.com/kassambara/survminer/issues/195
10. López-Ratón M, Rodríguez-Álvarez MX, Cadarso-Suárez C, et al. OptimalCutpoints: an R package for selecting optimal cutpoints in diagnostic tests. J Stat Softw 2014;61:1-36.
11. gtsummary. [Cited 2021 Mar 23]. Available online: http://www.danieldsjoberg.com/gtsummary/
12. Papageorgiou G, Grant SW, Takkenberg JJM, et al. Statistical primer: how to deal with missing data in scientific research? Interact Cardiovasc Thorac Surg 2018;27:153-8. [Crossref] [PubMed]
13. Leffondré K, Abrahamowicz M, Siemiatycki J, et al. Modeling smoking history: a comparison of different approaches. Am J Epidemiol 2002;156:813-23. [Crossref] [PubMed]
14. Motus I, Giss N. Tracheal stenosis. Prevention and treatment. Eur Respir J 2013;42:4955.
15. Motus I, Giss N, Medvinsky I, et al. Tracheal stenosis. One region experience. Eur Respir J 2021;58:PA426.
16. Motus IY, Konstantinova OS, Giss NA. Cicatricial tracheal stenosis. Is prevention possible? A single-region experience. Khirurgiia (Mosk) 2023;42-7. [Crossref] [PubMed]
17. Bacchin D, Aprile V, Lenzini A, et al. Surgical treatment of tracheal stenosis during Covid-19 era: a single-center experience and lessons learnt on the field. Updates Surg 2023;75:1681-90. [Crossref] [PubMed]
18. Wright CD, Li S, Geller AD, et al. Postintubation Tracheal Stenosis: Management and Results 1993 to 2017. Ann Thorac Surg 2019;108:1471-7. [Crossref] [PubMed]
19. Wright CD, Grillo HC, Wain JC, et al. Anastomotic complications after tracheal resection: prognostic factors and management. J Thorac Cardiovasc Surg 2004;128:731-9. [Crossref] [PubMed]
20. Lugg ST, Agostini PJ, Tikka T, et al. Long-term impact of developing a postoperative pulmonary complication after lung surgery. Thorax 2016;71:171-6. [Crossref] [PubMed]
21. Bibas BJ, Terra RM, Oliveira AL Junior, et al. Predictors for postoperative complications after tracheal resection. Ann Thorac Surg 2014;98:277-82. [Crossref] [PubMed]
22. Puri HV, Asaf BB, Mundale VV, et al. Predictors of Anastomotic Complications After Resection and Anastomosis for Tracheal Stenosis. Indian J Otolaryngol Head Neck Surg 2021;73:447-54. [Crossref] [PubMed]
23. Bohadana A, Teculescu D, Martinet Y. Mechanisms of chronic airway obstruction in smokers. Respir Med 2004;98:139-51. [Crossref] [PubMed]
24. Heijink IH, Brandenburg SM, Postma DS, et al. Cigarette smoke impairs airway epithelial barrier function and cell-cell contact recovery. Eur Respir J 2012;39:419-28. [Crossref] [PubMed]
25. Ravikumar N, Ho E, Wagh A, et al. The role of bronchoscopy in the multidisciplinary approach to benign tracheal stenosis. J Thorac Dis 2023;15:3998-4015. [Crossref] [PubMed]
26. Marchioni A, Andrisani D, Tonelli R, et al. Stenting versus balloon dilatation in patients with tracheal benign stenosis: The STROBE trial. Laryngoscope Investig Otolaryngol 2022;7:395-403. [Crossref] [PubMed]
27. Lavrysen E, Hens G, Delaere P, et al. Endoscopic Treatment of Idiopathic Subglottic Stenosis: A Systematic Review. Front Surg 2019;6:75. [Crossref] [PubMed]
28. Catano J, Uzunhan Y, Paule R, et al. Presentation, Diagnosis, and Management of Subglottic and Tracheal Stenosis During Systemic Inflammatory Diseases. Chest 2022;161:257-65. [Crossref] [PubMed]
29. Timman ST, Schoemaker C, Li WWL, et al. Functional outcome after (laryngo)tracheal resection and reconstruction for acquired benign (laryngo)tracheal stenosis. Ann Cardiothorac Surg 2018;7:227-36. [Crossref] [PubMed]