Broadening the scope for three-dimensional printed airway stents

To the Editor:

Currently, there is an arsenal of silicone, self-expandable metallic and biodegradable tracheobronchial stents that are available for the vast majority of malignant and benign central airway obstructions (CAOs) [1]. Unfortunately, airways stents are associated with significant complications, such as stent migration, mucus plugging and stent-edge granulation tissue development [2]. Prospective, systematic and objective evaluations of these complication rates are sparse and comparison between studies is hampered by the lack of uniform reporting. The quest for the ideal stent continues. Technical innovation in three-dimensional (3D) printing allows us to design and construct personalised airway stents that are congruent to the airway. This should result in optimal relief of the airway stenosis and improvement of symptoms in patients with CAO. We aimed to summarise the available literature on 3D-printed airway stents, list the advantages and disadvantages of personalised stents in comparison to conventional stents, and stress whether the advantages of 3D-printed airway stents outweigh the costs.

The design of a personalised airway stents is based on the patient's computed tomography (CT) scan [3]. Segmentation of the airways from the CT scan provides a virtual 3D model, offering a more detailed overview of the airway architecture that could allow the interventional pulmonologist to construct a 3D stent shape that optimally relieves the CAO. The creation of this 3D stent shape demands extensive input from an experienced interventional pulmonologist, who might have limited experience in designing 3D stents. The interventional pulmonologist will need to virtually correct the stenotic portion of the airways (tubular shape) while matching the patient's anatomy on the extremities based on the expected outcomes of the first steps of the procedure (balloon dilation) and of the radial force needed.

The ideal 3D airway stent is easy to insert and remove, remains in place, has fewer stent-related complications such as mucus plugging or granulation tissue formation; and it must assess airway patency [4]. Two aspects that must be in balance are shape and mechanical functionality. Designing a 3D-printed airway stent allows for a wider range of shapes with different inlet and outlet sizes, length, and varying diameters; however, mechanical properties require more attention. One important property is the radial force of the stent exerted on the airway wall, which will depend on a combination of the stent thickness and diameter, and hardness of the stent material. A balance must be preserved between this radial force and resistance to buckling, as it is important that the stent is flexible enough to fold when inserting it into the airway. Additional studs or other patient-specific antimigration structures can support the mechanical functionality.

As the interventional pulmonologist's input on the previously mentioned characteristics is required for designing personalised stents, a case-by-case approach is also required, and this can be time-consuming. The success of the creation process demands good communication between the physician and the manufacturer, and it also requires a significant level of experience from the parties involved. This experience-based practice will need to be supported by more fundamental research on the mechanical properties of airway stents if we want 3D stents to be more routinely used. Standardisation of 3D segmentation software or 3D printers will also be a requirement. More importantly, regulatory issues regarding 3D printing must be unravelled. The US Food and Drug Administration recently released a paper discussing the need to clarify how 3D-printed devices and software should be regulated, apart from the final medical device itself [5].

We performed a systematic review of the literature concerning personalised airway stents with a total of 434 unique articles. After removal of duplications and screening of the remaining articles by title and abstract (S. van Beelen and I. Wijma), 23 articles were reviewed entirely. Eventually, five articles were included in this review describing the virtual design and actual implantation of 15 patients with a 3D-printed airway stent in the trachea or main bronchi [3, 6–9]. Two unpublished cases from the Netherlands Cancer Institute (NKI) were also included.

The underlying causes of CAO were granulomatosis with polyangiitis (n/N=2/17, 11.8%), tracheobronchomalacia (n/N=4/17, 23.5%), post-transplant malacia and stenosis (n/N=5/17, 29.4%), post-tracheotomy complex stenosis (n/N=3/17, 17.6%), and bronchus stenosis after lobectomy or pneumonectomy (n/N=3/17, 17.6%). Most stents were Y-shaped and placed in the trachea/main bronchi (n/N=8/17, 47.1%), right main bronchus/lobar bronchi (n/N=3/17, 17.6%) and left main bronchus/lobar bronchi (n/N=3/17, 17.6%). The straight stents (n/N=3/17, 17.6%) were all placed in the trachea of which two migrated (cases 9 and 10). In total, eight stents (47.1%) were removed: six stents showed multiple complications (stenosis, mucus plugging, cough and/or migration), one stent was removed because placement was difficult, and one stent was removed per protocol. Wall thickness varied among and within stents between 1.0 and 2.0 mm (table 1). Some stents (n/N=5/17, 29.4%) had outer surfaces covered with studs, mostly depending on manufacturer or hospital preference.

To the best of our knowledge, this research is reporting the first review of 3D airway stents. Publications point out that personalised airway stents achieve good congruence to the airway. However, even if a stent perfectly fits the patient's airway anatomy, it remains a foreign body and complications can still occur. Mucus plugging, in particular, remained a concern in stents with different, nonconstant diameters and complex shapes. Straight stents, either 3D-printed or conventional, are prone to migration. Only if the anatomy is complex enough to anchor the stent to the airway can a 3D-designed stent be of benefit. Further research is needed to investigate variables such as the optimal stent material for each individual patient, how to optimise the radial force and hydrophilic properties without impairing the resistance to buckling, and whether it is possible to design better personalised stents based on a patients distorted airway. We hypothesise that 3D stents might be of great value if a complex shape stent needs to be designed (post-transplant anastomotic strictures or anastomotic stenosis following complex surgery such as sleeve lobectomy).

Several other limitations deserve to be mentioned. Studies and case series presented challenges in terms of comparison due to nonuniform ways to describe the stents and their outcomes. We, therefore, propose within this paper to describe the parameters mentioned in table 1 for each airway stent to allow a better comparison between publications with description of (mechanical) properties of the stents. Additionally, publication bias cannot be eluded, as many isolated cases are likely unpublished.

Although the effectivity of patient-specific airway stents has been proven in some publications [4, 7, 9], data remain scarce and long-term outcomes are still not reported. The TATUM phase II trial (www.clinicaltrials.gov identifier number NCT04848025), currently running in France (Toulouse and Marseille University Hospitals), which studies patient-specific stents for all planned airway stents with close follow-up, will provide long-term safety and efficacy data in a larger population. Further research is also required to investigate for which CAOs will personalised airway stents outperform conventional stents. A randomised controlled trial comparing conventional and personalised devices would be an optimal way to assess this. Indications for 3D stents are currently limited to rare cases (anatomically complex stenosis). In addition, conventional stents are readily available while 3D-printed stents require time to be manufactured. Evidence indicates that the ideal stent is not yet available; thus, the search for the ideal stent must continue.