1Dept of Pulmonary Medicine and Thoracic Surgery, Institut Universitaire de Cardiologie et de Pneumologie de Quebec, Quebec City, Quebec, Canada. 2Dept of Thoracic Oncology, Pleural Disease and Interventional Pulmonology, Hôpital Nord, Marseille, France.
Correspondence: Hervé Dutau, Dept of Thoracic Oncology, Pleural Disease and Interventional Pulmonology, Hôpital Nord, Chemin des Bourrely, 13015 Marseille, France. E-mail: email@example.com
In this chapter, we review the most frequent indications for airway stenting, the types of airway stents available and the data about potential new airway stenting technologies (biodegradable, drug-eluting and three-dimensional printed airway stents). We also discuss the complications associated with airway stenting and the necessary surveillance after airway stenting.
Cite as: Fortin M, Dutau H. Airway stents. In: Herth FJF, Shah PL, Gompelmann D, eds. Interventional Pulmonology (ERS Monograph). Sheffield, European Respiratory Society, 2017; pp. 236–251 [doi.org/10.1183/2312508X.10010117].
The first anecdotal reports of airway stent insertion date back as early as the 19th century [1, 2], but it was not until the 1960s when a dedicated airway stent became available and was widely used. The Montgomery T-tube, a silicone T-tube with an external side limb that protrudes through a tracheostomy, was the first available dedicated airway stent. Its use was limited to the upper trachea due to its side limb and method of insertion. Attempts to solve the problem of more distal airway stenosis were subsequently made by deploying modified vascular metal stents, which were unfortunately associated with unacceptable complications rates . In 1987, a revolution in airway stenting occurred with the introduction of Dumon stents (Novatech, La Ciotat, France), dedicated silicone stents for the trachea and bronchi . This innovation marks the beginning of interventional pulmonary medicine for many. These stents rapidly became widely used. Later in the 1990s, SEMS designed for the airways provided better clinical results than the previously used modified vascular stents and also gained in popularity . Many stents are now available with different characteristics. A significant quantity of data about airway stenting has been collected over recent decades. Here, we review the indications and complications for airway stenting, the types of stents currently available and their characteristics, as well as future avenues for airway stenting.
Malignant airway obstruction
Malignant airway obstruction (MAO) represents the most frequent indication for airway stenting (figure 1). Lung cancer is by far the most frequent cause of MAO; however, other cancers can directly invade or compress the airway (e.g. oesophageal and thyroid cancer or mediastinal lymphomas) and almost any cancer can cause distant metastasis to the airway. Patients present most frequently with dyspnoea, cough, symptoms of post-obstructive pneumonia and stridor.
MAO can be intrinsic, extrinsic or mixed. In cases of intrinsic stenosis, i.e. cases where the tumour has invaded the airway, the tumour should be debulked first. Once the tumour is debulked the clinician has to make the decision to stent the airway or not. In cases where significant residual stenosis is present, stenting is necessary. In cases where patency of the airway has been regained, the decision will depend on the expected response to upcoming therapies and the risk of stenosis recurrence. As an example, if complete patency of the airway has been obtained and the patient has a confirmed small cell lung cancer that will undergo rapid treatment, stent insertion is not required. In the opposite situation, where a patient has a nonsmall cell lung cancer and has already received multiple lines of chemotherapy and a maximal dose of radiation to the treated area, stenting is indicated due to the high risk of restenosis. A grey area between these two cases exists.
We participated in a randomised controlled trial of airway stenting versus no stenting in MAO . Airway stenting after relieving the obstruction significantly improved quality of life for a longer period of time over no airway stenting in patients with MAO, but the effect of stenting on quality of life was not statistically significant in patients that had not undergone any prior chemotherapy or radiation therapy. Except in occasional situations where it is used as a bridge to another therapy, such as chemotherapy or surgery, airway stenting in MAO is a palliative procedure and its role is to alleviate symptoms; hence, no intervention is required in asymptomatic patients.
Factors have been identified to help predict symptomatic response to airway stenting. There must be normal parenchyma distal to the obstruction. If atelectasis distal to the airway obstruction is present, traditional teaching has been that it should not have been present for many months as surfactant becomes dysfunctional and parenchyma cannot re-expand. Experience from endobronchial valves for lung volume reduction can lead us to reconsider this as atelectasis has been demonstrated to resolve after being present for prolonged periods. The time after which lung parenchyma will not re-expand even if the occluded airway is reopened is not clear. Perfusion to the parenchyma to which the obstructed airway leads must be sufficient. Hypoxic vasoconstriction caused by decreased ventilation in MAO may improve after alleviating the airway obstruction; however, if perfusion to the concerned area is absent due to vascular obstruction by the malignant tissue, perfusion will not improve after relieving the airway obstruction . As an example, if the right mainstem bronchus is completely occluded as well as the right pulmonary artery, the patient will not benefit from regaining ventilation to their right lung and might in fact become more symptomatic as a ventilatory dead space will be created when the right mainstem bronchus obstruction is relieved. The benefit of relieving airway obstruction that is lobar or beyond is also not clear. The MAO portion of the ACQuIRE cohort, a large prospective cohort of patients undergoing interventional pulmonary procedures in North America, showed that lobar obstruction or more distal obstruction was a predictor of the absence of symptomatic response to procedures . Patency of distal smaller airways can be obtained during a procedure, but MAO has a high probability of rapid recurrence if these small airways are surrounded by a tumour and are too small to be stented. No specific stent has proven its superiority in this indication.
Patients with MAO generally have limited survival and the long-term complications of stenting are less of a concern in this population, although this could change with the recent appearance on the market of medications that may significantly prolong the survival of lung cancer patients.
Benign airway stenosis
Idiopathic, post-intubation, post-tracheotomy, post-lung transplantation and surgical anastomotic airway stenosis are the most frequent causes of benign airway stenosis (BAS). Less frequent causes include papillomatosis, amyloidosis, granulomatosis with polyangiitis, sarcoidosis and tuberculosis. The clinical presentation of patients with BAS is similar to patients with MAO, but the course is generally more insidious and hence symptoms are more often long-standing.
It is important to distinguish the management of BAS from the management of MAO as patients with BAS tend to have a prolonged survival that will expose them to long-term complications of airway stenting. Stenting should not be considered as standard first-line therapy in BAS, as surgery is the treatment of choice, but as an alternative therapy when a patient is not a good surgical candidate for technical reasons or because of comorbidities. As an example, granulation tissue has more time to form and become symptomatic or metal has more time to fatigue and break in BAS than in MAO. A BAS will also tend to be stiffer and more irregular in shape, giving less of an opportunity for a stent to “anchor” in position.
Distinguishing simple and complex stenosis is also important in the approach to the endoscopic treatment of BAS. Complex stenosis is defined by the presence of cartilaginous involvement . Simple or web-like stenosis, such as often seen in idiopathic tracheal stenosis, will frequently not recur after endoscopic treatment without stenting [10, 11]; hence, stenting should not be a first-intention treatment in this population, but it can be used in stenosis refractory to endoscopic treatment without stenting or complex BAS.
Endoscopic success rates using silicone stents can reach 70% in benign tracheal stenosis (figure 2)  and in post-lung transplant anastomotic stenosis . One must always consider surgical resection of the stenosis before stenting as this is the therapy of choice for the vast majority of BAS. Cases of BAS should be discussed with a surgeon with expertise in tracheal resection. Reports of treatment of stenoses with stents causing extension of the disease by stent-related complications that precluded future surgeries have been published . One should be cautious not to cause further strictures with airway instrumentation. For this reason, partially covered SEMS are contraindicated in BAS and benign airway disease.
Patients with EDAC may present with dyspnoea, cough or recurrent infections. Stenting has resulted in mitigated results in this population in our experience. ERNST et al.  confirmed this impression in an interesting study on tracheobronchomalacia, demonstrating that respiratory symptoms can be improved in the short term, but that short- and long-term complications are frequent. Further data are required to determine the role of stenting in EDAC, but the preliminary data do not point towards airway stenting being a long-term solution. Others have used airway stenting to predict clinical response to surgical correction of EDAC . Stents are left in place for a short period of time to evaluate the symptomatic response of patients and, if symptoms improve sufficiently, patients are deemed potential responders for a surgical procedure.
Tracheo-oesophageal fistulas (TOFs) are a cause of recurrent pulmonary infections. They can be benign (endotracheal intubation, tracheostomy or tuberculosis) or malignant (pulmonary or oesophageal cancer) in origin. The role of stenting in TOFs is mainly in malignant cases and, occasionally, in benign cases in patients unfit for surgery. Although all agree that malignant TOFs should be closed, it is unclear if this should be done with an oesophageal stent, a tracheal stent or both. When treating a malignant TOF with an oesophageal stent, clinicians should be aware of the risk of compromising the airway by pushing the tumour towards the airway lumen. For this reason, patients with symptomatic tracheal and oesophageal obstructions should all have an airway stent inserted prior to oesophageal stent insertion . For patients without symptomatic airway obstruction, we prefer to start with oesophageal stenting and follow-up with the patient to see if a fistula persists and if respiratory symptoms develop. If the fistula persists despite an optimally positioned oesophageal stent or if symptomatic airway compression develops after airway stenting, an airway stent can be inserted. Fistulas are discussed in more detail elsewhere in this Monograph .
Surgical dehiscence and airway lacerations in patients unfit to undergo surgical repair may also be treated with airway stenting . Airway lacerations in patients not requiring positive pressure ventilation will heal spontaneously and do not require stenting.
Type of stents
Polymer versus metal stents
In the early 1990s the classification of stents was simple as only Dumon silicone stents (figure 3), Montgomery silicone T-tubes and first-generation metal stents existed. Since then the first-generation metal stents (e.g. Gianturco (Cook Medical, Bloomington, IN, USA) and Palmaz (Johnson & Johnson, Somerville, NJ, USA) stents) have become obsolete because of their high complication rates [13, 19, 20].
To solve the problems associated with the erosion of metal stents in the airways, partially covered second-generation SEMS were developed. These stents are partially covered with polymer, but their metal frame remains bare in the distal portions to theoretically allow the stent to anchor into the mucosa and prevent migration. These dedicated airway stents provided better clinical results [5, 21, 22], although significant complications remained when they were left in place for prolonged periods of time, as became evident after several years of use in patients with BAS and prolonged survival [23, 24]. This led to a “black box warning” from the US Food and Drug Administration for the use of SEMS in BAS in 2005 .
Third-generation SEMS were introduced to overcome these problems. These are SEMS which are fully covered with polymer and which theoretically should not present the same long-term complications as partially covered second-generation SEMS. Little data is available about these stents, although they have been commercialised for several years. Data about their efficacy and safety in MAO was initially published by MARCHESE et al.  and we recently published our experience with a third-generation SEMS in BAS, demonstrating an acceptable rate of minor complications that were all managed endoscopically with success . Migration rates were high in both studies at 13.4% and 32.5%, respectively.
To the best of our knowledge, polymer stents and SEMS have never been directly compared. Efficacy and complications associated with a stent will vary depending on the indication for its use and the duration it is left in place; hence, it is difficult to compare stents by comparing the results of different studies in which the populations differ. As an example, the same stent model was used in our recent series  of BAS and the MARCHESE et al.  series of MAO, but our complication rate differed as our patients had benign disease and lived significantly longer with their stents in place. Polymer stents have the advantage of being customisable at the bedside (figure 4) . As an example, if a MAO involves the proximal centimetre of the right mainstem, the distal centimetre of the trachea and the proximal centimetre of the left mainstem, a Y-stent can be customised at the bedside to fit these exact dimensions after measurements are taken endoscopically , whereas a one-size-fits-all stent will have to be selected if a SEMS is used, which in the end will be unnecessarily large. Customising polymer stents will also allow treatment options that are not possible with SEMS, such as creating a window in the stent to ventilate a right upper lobe. SEMS have the advantage of not requiring rigid bronchoscopy placement. In our opinion, this is not a real advantage as an individual placing an airway stent should be competent in rigid bronchoscopy, and should be able to convert to rigid bronchoscopy if a complication occurs and this technique is required.
SEMS are available in larger diameters (figure 5) than polymer stents as larger polymer stents cannot be folded to a small enough diameter to fit down a rigid bronchoscope and be positioned . This difference may be of clinical interest in the treatment of tracheal disease where the largest diameter of silicone stent commercially available may be insufficient, increasing the risk of migration. Our previous work has shown a lower risk of migration with Silmet stents (Novatech, La Ciotat, France) than with Dumon silicone stents in dynamic A-shape tracheal stenosis, a situation where large tracheal stents are needed . SEMS also have a thinner profile that leaves a larger residual lumen for the same outside diameter, which may facilitate mucus clearance. We will not review all stents available on the market here, as this could be the topic of a full chapter in itself, but we refer the reader to previous work from FREITAG  and our group  for detailed information on available airway stents.
As already noted, no clinical trial has compared different stents head to head. Only data comparing mechanical properties of different stents are available [33, 34]. Different stents have different characteristics and no stent is perfect; hence, clinical judgement should be used to choose the optimal airway stent for an individual patient. A recent survey of members of European Association for Bronchology and Interventional Pulmonology demonstrated that a wide variety of airway stents are used across Europe . Dumon silicone stents and Ultraflex SEMS (Boston Scientific, Marlborough, MA, USA) (figure 6) were the most commonly used airway stents.