Chapter 15
Central airway obstruction
Christophe Dooms1 and Antoni Rosell2
1Dept of Respiratory Diseases, University Hospitals KU Leuven, Leuven, Belgium. 2Bronchoscopy Unit, Dept of Respiratory Medicine, Hospital Universitari de Bellvitge, Universitat de Barcelona, IDIBELL, L’Hospitalet de Llobregat, Barcelona, Spain.
Correspondence: Christophe Dooms, Dept of Respiratory Diseases, University Hospitals KU Leuven, UZ Leuven Campus Gasthuisberg, Herestraat 49, 3000 Leuven, Belgium. E-mail: christophe.dooms@uzleuven.be
Neoplastic central airway obstruction (CAO) with imminent respiratory failure, stridor and/or severe dyspnoea requires immediate and appropriate care and action. Initial evaluation of CAO involves CT and bronchoscopy in order to decide upon timely referral for interventional pulmonology, consisting of mechanical debulking with or without thermocoagulation to restore airway patency and with or without airway stenting to preserve central airway patency. The assessment and treatment of benign central airway strictures should be reserved for selected centres that provide a multidisciplinary, dedicated interventional approach and are evaluated based on qualitative long-term outcome.
Cite as: Dooms C, Rosell A. Central airway obstruction. In: Herth FJF, Shah PL, Gompelmann D, eds. Interventional Pulmonology (ERS Monograph). Sheffield, European Respiratory Society, 2017; pp. 224–235 [https://doi.org/10.1183/2312508X.10003717].
Patients with central airway obstruction (CAO) often have nonspecific clinical manifestations that are responsible for a delayed diagnosis. The reason for this delayed diagnosis, but often acute presentation, is that symptoms usually begin at 50% airway narrowing and from there airflow resistance increases exponentially as the stenosis becomes more severe. Significant CAO generally presents with severe dyspnoea, stridor, orthopnoea and/or imminent respiratory failure requiring immediate action [1].
How to evaluate a CAO
The central airways are defined as those airways that can be directly visualised by flexible therapeutic bronchoscopy with an outer diameter of 5–6 mm. The central airways of interest for interventional pulmonology are limited to the trachea, mainstem bronchi and lobar bronchi.
CAO can be classified into two distinct patterns, i.e. focal airway obstruction and diffuse airway narrowing, which are generally split into malignant and nonmalignant causes of disease (table 1). CT and bronchoscopy are complimentary techniques that play an important role in the assessment, diagnosis and treatment planning for a CAO.
Table 1. Classification of central airway obstructions into two distinct patterns
Focal tracheobronchial airway narrowing |
Post-traumatic stricture: post-intubation stenosis, post-tracheotomy stenosis, postoperative airway anastomotic stricture (e.g. lung transplantation, sleeve lobectomy) |
Post-infectious stricture (e.g. tuberculosis, Aspergillus, rhinoscleroma) |
Neoplasia: benign neoplasia (e.g. hamartoma, papilloma, lipoma, etc.), primary tracheobronchial carcinoma, metastatic disease to tracheobronchial wall, mediastinal neoplasia with direct tracheobronchial invasion |
Inflammatory disease with or without autoimmune disorder: idiopathic subglottic tracheal stenosis, granulomatosis with polyangiitis, rheumatoid arthritis, inflammatory bowel disease, IgG4-related disease, sarcoidosis, amyloidosis |
Extrinsic compression (e.g. thyroid tissue, mediastinal adenopathy, vascular structure) |
Diffuse tracheobronchial narrowing |
Post-traumatic stricture: inhalation injury |
Post-infectious stricture (e.g. tuberculosis, Aspergillus, etc.) |
Neoplasia: malignant or benign (e.g. papilloma) |
Inflammatory disease with or without autoimmune disorder: granulomatosis with polyangiitis, sarcoidosis, relapsing polychondritis, inflammatory bowel disease, amyloidosis, tracheobronchopathia osteochondroplastica |
Dynamic tracheobronchomalacia and expiratory dynamic airway collapse |
Contrast-enhanced multidetector CT in deep inspiration acquires 1–5 mm sections throughout the thorax for routine two-dimensional axial, coronal and sagittal multiplanar reconstruction. It also provides a variety of three-dimensional techniques for optimal imaging of the central airways to enhance estimation of the type, length and degree of stenosis, along with visualising the airways distal to the point of obstruction and/or the relationships between central airways and adjacent mediastinal and hilar structures (e.g. blood vessels and lymph nodes). Other respiratory disorders can also be detected, such as bronchiectasis, pulmonary embolism and parenchymal disorders. Limitations of the CT scan include the inability to reliably differentiate mucosal from submucosal or extrinsic disease, inability to visualise subtle (sub)mucosal extension, inability to recognise mucus impaction around a CAO and an often inaccurate evaluation of airway patency distal to an occluding CAO. A CT scan might not be possible in some cases of severe orthopnoea where the patient cannot lay flat.
Video bronchoscopy is generally more accurate in assessing the exact location of a tumour (distance from larynx or carina as well as distance assessment in order to evaluate stent compatibility), the degree, length and type of airway stenosis, and distal airway patency. It also has the advantage of facilitating a biopsy for histopathological diagnosis. Video bronchoscopy is the preferred tool to assess whether the CAO consists of intraluminal tumour growth, extraluminal tumour compression or a combination of both. Additionally, tracheomalacia and excessive dynamic central airway collapse are better defined during bronchoscopy.
An excellent standardised qualitative classification scheme designed for grading tracheobronchial stenosis from the pulmonologist’s perspective, taking into account the CT scan and bronchoscopy evaluation, was proposed in 2007 with descriptive images and diagrams for rapid and uniform classification of types of central airway stenosis [2]. We recommend its use in daily clinical practice, although the degree of severity criterion is not justified physiologically. There is currently no commonly accepted standardised technique to quantify the severity or degree (percentage cross-sectional area) of large airway luminal narrowing by bronchoscopic inspection [3]. Preferred techniques vary considerably and are often based on a subjective visual estimation that is poorly reproducible. Using a classification scheme of mild (<50%), moderate (50–70%) and severe (>70%) obstruction to evaluate still bronchoscopic images resulted in 53% incorrectly classified stenoses, of which 9% and 91% were overclassified and underclassified, respectively [4]. Relying upon the degree of obstruction and not taking into account symptoms and patient comorbidity can significantly affect patient care: underestimation can lead to inadequate or delayed therapy, while overestimation may lead to referral for unnecessary invasive interventions.
Studies of virtual bronchoscopy have found fairly good correlation between the degree of endoluminal airway narrowing noted on virtual imaging and bronchoscopy. Virtual bronchoscopy is based on CT scans with 1–1.5 mm sections together with the use of CT segmentation and three-dimensional volume rendering. Advances in real-time objective measurements from video bronchoscopy are also being explored [5]. Finally, a perfect measurement technique should be integrated in each particular clinical scenario with malignant CAO, as it is equally important to consider the functional status of the larynx and small airways, as well as body mass index and other comorbid limitations, when assessing the need for a central airway bronchoscopic treatment.
General algorithm and impact of malignant airway recanalisation
When a patient with suspected significant upper airway obstruction presents acutely at the emergency room, selective supportive measures may be necessary and include the commencement of an inspired helium–oxygen (Heliox) mixture. Heliox is less dense than oxygen and has the effect of predisposing to laminar flow, which can be used to improve airway dynamics in the short term. Securing the airway for adequate oxygenation and ventilation must be guaranteed in the emergency room. This can be achieved by laryngeal mask airway ventilation, by passing an ETT beyond the central airway stenosis or placing the tip of the ETT above the growth while maintaining ventilation, or by referring the patient for urgent rigid bronchoscopy and removal of the tumour. An urgent tracheotomy is generally only considered in clinical cases already evaluated by bronchoscopy and considered for tracheotomy but not for rigid bronchoscopy.
When a patient presents with an upper airway obstruction in a nonacute setting, an elective assessment and management plan should be proposed. Interventional pulmonology as part of this plan must be carefully considered and executed to obtain an optimal outcome, i.e. improvement in quality of life and survival [6–9]. A typical interventional pulmonology algorithm for central airway recanalisation encompasses the following five phases:
1)Query whether a symptomatic stenosis is present and/or palliation is required.
2)Identify the cause, morphology, extent and mechanism of the central airway narrowing, together with confirmation of patent distal airways (figures 1a and b, 2a and b, and 3a–c).
3)Treat the intraluminal component when present (figures 1c and d, and 3d).
4)Assess the residual degree of airway narrowing (figures 1e and 2c).
5)Treat the residual stenosis and/or mural component when considered appropriate (figure 2d).
A multicentre national registry in the USA reported a technical success rate to re-open a >50% narrowing of the airway lumen in 93% (range 90–98%) of patients, while a clinically significant improvement was only obtained in 42% and 48% of patients for quality of life and symptoms such as dyspnoea, respectively [10]. Two prospective single-centre cohorts have been published [11, 12]. In both cohorts, the interventional pulmonology procedure was considered successful in 91–92% of patients as a post-intervention lumen patency of at least 50% was achieved for a pre-intervention symptomatic >50% narrowing of the central airway. In one cohort it was reported that a completely restored airway patency (defined as 80–100% patency) was obtained in 41% of patients, while a partially restored airway patency (defined as 50–80% patency) was obtained in 50% of patients [12]. The later might be relevant in order to explain the lower rate of subjective benefits reported. OVIATT et al. [11] reported a minimal clinical improvement of 50 m in 6-min walk distance at 30 days in 32% of patients and a minimal clinical improvement in composite dyspnoea scores by European Organisation for the Research and Treatment of Cancer quality of life questionnaires (core QLQ-C30 and lung cancer QLQ-LC13 modules) in 46–49% of patients at 1 month. STRATAKOS et al. [12] compared a group of patients with a symptomatic >50% CAO who declined an endoscopic interventional pulmonology treatment (control group) to a group who underwent a bronchoscopic interventional pulmonology treatment (intervention group). Quality of life significantly improved and dyspnoea decreased in the intervention group, not deteriorating for those who survived up to 12 months post-procedure, whereas patients who declined an interventional treatment (control group) had worse quality of life and dyspnoea at all follow-up time-points [12].
Figure 1. Case illustration of an algorithm for malignant right main bronchus obstruction. a) Coronal CT scan section showing a tumour originating in the ostium of the right upper lobe with right mainstem occlusion. b) Bronchoscopy showing an occlusion of the distal right main bronchus and signs of recent bleeding, demonstrating a squamous cell carcinoma at histopathology. c) Bronchoscopic Nd-YAG laser coagulation of the occluding tumour in order to obtain haemostasis. d) Bronchoscopic status after Nd-YAG laser vapourisation. e) Final status after endoscopic debulking with thermocoagulation showing patent distal airways.
