3D printing as an innovative tool in personalized management of complex airway diseases: a literature review

Abstract

Background and Objective

Personalized medicine tailors interventions to a patient’s unique anatomy and physiology. Three-dimensional printing (3DP) enables this precision for complex airway disease, including tracheal stenosis, tracheobronchomalacia, aerodigestive fistulas, and segmental defects, where conventional silicone or metallic stents and surgical reconstruction often fail to provide durable, anatomically congruent solutions. Tissue engineering and 3DP promise patient-specific devices and regenerative scaffolds that maintain patency, resist collapse, and minimize immunogenicity. This review synthesizes clinical and preclinical progress, highlighting materials, design strategies, biologic integration, and translational barriers.

Methods

A comprehensive literature search was conducted in PubMed (January 1, 2015–June 1, 2025). Inclusion criteria encompassed studies utilizing 3DP to fabricate implantable devices for tracheobronchial reconstruction, with in vivo implantation. Pediatric (<18 years), egg/mouse/rat preclinical studies, review articles, and abstracts were excluded. Data extracted included publication details, participant characteristics, device materials and printing methods, and outcomes.

Key Content and Findings

From 808 records, 16 clinical and 56 preclinical studies were analyzed. Clinically, indirect 3DP with silicone or metallic alloys predominated, creating Y-stents or straight stents for post-lung transplant (LTx) stenosis, tracheobronchomalacia, granulomatosis with polyangiitis, malignant obstruction, and aerodigestive fistulas. 3DP technologies facilitate the synthesis of customized stents that can better conform to individual airway geometries, offering more precise therapeutic options than conventional one-size-fits-all devices. In parallel, preclinical studies aim to address the limitations observed within clinical settings by focusing on long-term, regenerative solutions. Preclinical studies focused on biodegradable scaffolds, commonly polycaprolactone (PCL), enhanced through surface modification or hybridization with hydrogels such as gelatin methacryloyl (GelMA) or silk fibroin and bioactive factors like transforming growth factor-β (TGF-β) or stromal cell-derived factor-1 (SDF-1). Bilayer constructs with epithelial and chondrogenic components supported epithelialization, cartilage formation, and vascularization. Advanced strategies such as exosome use, ferroptosis inhibition, and heterotopic preconditioning improved integration.

Conclusions

3DP enables anatomically tailored airway implants and promising regenerative scaffolds. Translation is limited by technical variability, regulatory complexity, and sparse long-term data. Standardized protocols, rigorous trials, and multidisciplinary collaboration are essential to bring 3DP airway reconstruction into clinical practice.

Introduction

Personalized medicine seeks to deliver patient-specific interventions by considering an individual’s unique anatomical and physiological profile. Many technologies have emerged to enable this paradigm shift from the conventional “one size fits all” approach to precision medicine, with three-dimensional printing (3DP) at the forefront (1,2). In the context of complex airway diseases, 3DP offers potential to overcome significant therapeutic challenges inherent to the complex and dynamic anatomy of the tracheobronchial tree. Diseases, such as tracheal stenosis, tracheomalacia, aerodigestive fistulas, and segmental airway defects, are difficult to manage and carry high rates of morbidity and mortality (3). Current minimally invasive treatments may involve commercially available stents, typically made from silicone/metallic alloys, to restore luminal patency and provide symptomatic relief. In more complex cases, surgical reconstruction may be required, often through resection with end-to-end anastomosis; however, in adults with long-segment tracheal defects greater than half the total tracheal length, end-to-end anastomosis is not possible (4).

In these cases, treatment typically relies on various tracheobronchial replacement strategies, each with their own risks and benefits, including synthetic prostheses, tracheal or aortic allografts, autologous grafts, and tracheal transplantation (4-6). In recent years, cryopreserved aortic allografts have become the most widely used method for tracheobronchial replacement, particularly for extensive malignant lesions (5). However, non-tissue engineering approaches may be constrained by donor site morbidity, immunosuppression, and graft failure (7). Tissue-engineered constructs seek to address some of these limitations, primarily using decellularized human trachea and autologous cells, synthetic scaffolds and autologous cells, or the use of in vivo bioreactors (5). For example, Martinod et al. used stented cryopreserved aortic matrices as an in vivo bioreactor for patients with end-stage tracheal lesions or proximal lung tumors and demonstrated low 90-day mortality, feasibility of stent removal, and evidence of in vivo epithelial and cartilage regeneration (8).

Despite available options, these existing treatments often fall short of achieving durable, long-term airway reconstruction for both short- and long-segment defects. This is particularly true in cases involving circumferential involvement or dynamic airway collapse. Although commercial airway stents can restore luminal patency in short-segment defects, they often result in suboptimal congruence and are associated with challenges related to airflow, secretion clearance, and high risks of migration, granulation, and tissue injury (9,10). Similarly, a recent systematic review by Tseng et al. highlighted that tracheal replacement techniques continue to produce variable outcomes due to limitations in mechanical stability and long-term integration (4). Thus, successful tracheal reconstruction must meet several requirements: provide mechanical strength to withstand respiratory forces, maintain airway patency, resist collapse, and minimize immunogenicity. Given these demanding criteria, there is a growing interest in personalized three-dimensional (3D)-printed devices.

Current research on 3D-printed airway solutions can be broadly categorized into two domains: clinical applications, which predominantly focus on removable metallic and silicone stents, and preclinical studies, which center on the development of biodegradable, tissue-engineered scaffolds intended for long-term integration. The emphasis on tissue regeneration in preclinical models stems from the limitations of current clinical devices, which often provoke complications such as migration, granulation, mechanical instability, and poor tissue integration (11). As such, researchers have turned to 3DP in combination with biomaterials to develop patient-specific constructs that can eventually eliminate the need for synthetic stents. Together, these parallel avenues of research reflect both the short-term clinical need for personalized airway stents and the long-term vision of achieving tracheal regeneration through 3D-printed bioengineered constructs. This review offers a comprehensive overview of the clinical applications of 3DP for complex airway reconstruction, along with an in-depth analysis of preclinical investigations, focusing on animal models, biomaterials, and strategies to enhance epithelialization and vascularization. By examining the progress in both domains, we aim to highlight persistent challenges and key areas for improvement to support safe and effective translation from the laboratory to clinical settings. We present this article in accordance with the Narrative Review reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-aw-2057/rc).

Methods

A comprehensive review was conducted using the PubMed database (Table 1). Results were restricted to studies published between January 1, 2015 and June 1, 2025. The year 2015 was selected as the starting point because the first clinically implanted 3D-printed airway stents were introduced around this timepoint. In parallel, given the breadth of preclinical investigations, restricting the review to this timeframe aimed to ensure that the included studies reflected more contemporary techniques with higher translational potential. Articles were included if they were original full-text articles in English that utilized 3DP to construct a tracheobronchial device for in vivo implantation into animals larger than rodents. Manuscripts were excluded if they (I) did not utilize 3DP to construct an implantable device; (II) were not focused on reconstruction in the tracheobronchial tree; and (III) did not perform implantation of the device in vivo. Clinical studies involving patients less than 18 years of age were excluded to maintain a uniform adult cohort given that pediatric airway disease often differs from adult pathology in anatomy, disease mechanisms, and management. Preclinical studies that involved implantation exclusively into eggs, mice, or rats were excluded. Additionally, any review articles, book chapters, erratums, and abstracts were excluded.

Table 1. The search strategy summary.

Items	Specification
Date of search	June 2, 2025
Database searched	PubMed
Search terms used	((three-dimensional) OR (3-D) OR (3D) OR (3-dimensional)) AND ((trachea) OR (tracheal) OR (airway)) AND ((print) OR (scaffold) OR (stent))
	Filter: 2015–2025
Timeframe	January 1, 2015–June 1, 2025
Inclusion and exclusion criteria	Inclusion: (I) utilized 3DP to construct an implantable device for tracheobronchial tree; (II) implanted device in vivo into species larger than rodents; (III) clinical studies involving patients older than 18 years of age; (IV) clinical case reports, case series, cohort studies, case-control studies, clinical trials, preclinical studies; and (V) full text available in English
	Exclusion: (I) did not utilize 3DP to construct an implantable device; (II) were not focused on reconstruction in the tracheobronchial tree; (III) did not perform implantation of the device in vivo; (IV) clinical studies involving patients less than 18 years of age; (V) preclinical studies that involved implantation exclusively into eggs, mice, or rats; (VI) review articles, book chapters, erratums, and abstracts; (VII) full text not available; and (VIII) non-English studies
Selection process	Performed by two independent reviewers (S.E.M. and N.J.I.) with high concordance. Any uncertainty was resolved through group discussion (S.E.M, E.A.T., S.W, and N.J.I.), and if necessary, a fourth reviewer was consulted (D.H.B.)

3DP, three-dimensional printing.

Data collected from each article included information pertaining to the (I) publication [title, author(s), journal, year of publication]; (II) participants (number, species, clinical diagnoses); (III) 3DP implantable device characteristics (material utilized, type of device, mechanism of print); and (IV) outcomes. Direct 3DP refers to the process in which the final scaffold or implantable device is fabricated directly from a digital model with or without post-processing steps. Indirect 3DP was defined as the creation of a mold or model using 3DP that is then used to create a unique implantable device via bench-top injection molding and shaping of metal/polymers, respectively, to render an end-use part. Commercial stents that were merely modified based on 3D-printed models were excluded. The corresponding authors of included clinical studies were contacted via email for additional details regarding the site of 3DP, manufacturing time, and time to device implantation, if the information was not provided in the published text.

Results

Study identification and selection

Our search generated 808 articles. Ultimately, 16 clinical and 56 preclinical studies met the inclusion criteria and were analyzed (Figure 1).

Figure 1.

Flow chart of the screening and selection process. 3DP, three-dimensional printing.

Personalized airway solutions: clinical applications of 3DP

Types of 3D-printed solutions

Since 2015, there have been more than a dozen clinical publications encompassing 86 patients (Table 2). Most used an indirect 3DP technique based on models created from computer tomography or magnetic resonance imaging (13,22,25). After generation of the model, a medical-grade material (typically silicone) is injected into the printed airway mold (9,13,15). In these studies (n=6, 37.5%), the stent manufacturing ranged from five to 30 days (9,15). Alternatively, in some cases (n=5, 31.3%), a flexible metal alloy was shaped around the 3D-printed model to achieve the desired configuration (20,21,24-26). Among these cases, manufacturing time ranged from 2 to 4 days. Only three studies (18.8%) utilized direct 3DP involving materials other than metals or silicone [carbon fiber, polyurethane, or polycaprolactone (PCL)] (14,18,19). Of these, stent manufacturing time was only available for one study, which was 10 days for a carbon fiber stent (19).

Table 2. Summary of clinical studies and outcomes.

Etiology	Number	Disease type	3DP technique	Material & device type	Outcome	Complications	Mean follow-up
Benign
Krumm and Gesthalter (12)	2	Extrinsic vascular compression, airway dehiscence post-LTx	Unknown	Silicone, Y-stent	The procedure was tolerated well, with near complete resolution of the air leak following stent placement	None reported	6 months
Sawal et al. (13)	9	Anastomosis stenosis in post-LTx	Indirect, injection molding	Silicone, Y-stent	Mean percent increase of 51.05% in FEV1 at 90 days. Able to remove stents in 2 patients	Granulation, n=2. Sputum retention, n=3. Stent migration, n=1	6 months
Annangi et al. (14)	1	Anastomosis stenosis post-LTx	Direct	Polyurethane, Y-stent	FEV1 improved from 41% to 84% and remained stable during follow-up period	Mucous plugging n=1 (due to patient nonadherence to mucolytic meds)	15 months
Guibert et al. (15)	10	Anastomosis stenosis, TBM	Indirect, injection molding	Silicone, Y-stent and straight stent	Patient-reported improvement in quality of life. Mean percent increase of 21.61% in FEV1 and 92.54% in PEFR at 90 days. 9/10 patients experienced decrease in NYHA classification	Stent migration, n=2. Mucus plugging, n=1. Total stent removal due to intolerance, n=3	4.5 months
Gildea et al. (9)	2	Tracheobronchial granulomatosis	Indirect, injection molding	Silicone, Y-stent and straight stent	Patient-reported improvement in quality of life. Deceased emergency/unplanned bronchoscopy	Granulation, n=1. Mucus plugging, n=1	12.7 months
Schweiger et al. (16)	2	TBM	Indirect, injection molding	Silicone, Y-stent	Patient-reported improvement QOL	None reported	6.5 months
Guibert et al. (17)	1	Anastomosis dehiscence & stenosis post-LTx	Indirect, injection molding	Silicone, Y-stent	Patient-reported improvement QOL. PEFR increased from 2.27 to 3.78 L/min 7 days after the procedure	None reported	2.5 months
Huang et al. (18)	1	TBM	Direct	PCL, suspension scaffold	Improvement of breathing and physical strength. Scaffold remained stable in mediastinum on imaging	None in follow-up period	3 months
Malignant
Wang et al. (19)	1	Large trachea defect repair	Direct	Carbon fiber, Straight	Symptom free after 8 weeks. No stenosis or obstruction during 3-year follow-up	None reported	3 years
Shan et al. (20)	10	Laryngotracheal stenosis	Indirect, shaping	Nickel titanium alloy with PTFE junction, Straight stent	Improved dyspnea. Significantly lower HJ classifications after the procedure	Granulation, n=2. Stent migration, n=1. Stent intolerance, n=1	4.9 months
Shan et al. (21)	12	Tracheobronchial stenosis	Indirect, shaping	Nickel titanium alloy with PTFE junction, Y-stent	Improved dyspnea. Significant improvement in HJ classification. Significant improvement in KPS/functional status	Granulation, n=2. Sputum retention, n=4	5.6 months
Ojima and Kitamura (22)	1	Adenocarcinoma compression	Indirect, molding	Silicone, Y-stent	Patient died 11 months after stent placement due to cancer	None reported	11 months
TEF
Smesseim et al. (23)	1	Post-esophagectomy malignancy	Unknown	Silicone, Y-stent	No granulation tissue seen at time of stent removal	Stent could not close the defect 2 months later	5 months
Shan et al. (24)	26	Post-esophagectomy malignancy	Indirect, shaping	Nickel titanium alloy with PTFE junction covered in silicone, Y-stent	Improvement in KPS/functional status	Granulation, n=2. Sputum retention, n=5. tent migration, n=1. Stent intolerance, n=1	4.9 months
Huang et al. (25)	6	Post-esophagectomy	Indirect, shaping	Nickel titanium alloy with PTFE junction, Y-stent	Significant improvement in KPS/functional status. Significant improvement in dysphagia	Granulation, n=1. Sputum retention, n=2	16.5 months
Han et al. (26)	1	Post-esophagectomy	Indirect	Nickel titanium alloy covered in silicone, Y-stent	Improved cough. Shrinkage of fistula (which was then able to be fully closed with injection of fibrin glue)	None reported	11.1 months

3DP, three-dimensional printing; FEV1, forced expiratory volume in 1 second; HJ, Hugh-Jones; LTx, lung transplant; KPS, Karnofsky performance status; NYHA, New York Heart Association; PCL, polycaprolactone; PEFR, peak expiratory flow rate; PTFE, polytetrafluoroethylene; QOL, quality of life; TBM, tracheobronchomalacia; TEF, tracheoesophageal fistula.

Materials and designs

Intraluminal 3D-printed stents reported in the clinical literature were designed as either Y-shaped or straight stents. Y-stents are typically indicated for use in cases involving the tracheal carina or mainstem bronchi bifurcations, offering support across multiple branches and accommodating dynamic angles (27). Straight stents, on the other hand, tend to be applied to lesions confined to the trachea or a single bronchus. Additionally, the use of 3DP to generate an external suspension scaffold rather than a luminal stent was reported, indicating the potential of 3DP technology to expand beyond intraluminal devices toward external structural support for airway stabilization (18). The most used materials are medical-grade silicone and metal alloys, due to their established safety profiles and mechanical reliability. Silicone offers flexibility, ease of shaping, and an established clinical track record in airway stenting (9,12,13,15-17,22,23). However, its non-degradable nature raises concerns about long-term complications such as migration, granulation tissue formation, and mucosal irritation (9,13,15). Furthermore, implantation of silicone stents requires rigid bronchoscopy (28). In contrast, metallic devices can be implanted with fiberoptic bronchoscopy and provide excellent radial force and structural integrity, making them well-suited for dynamic airway environments (20,21,24-26). However, compared to silicone stents, metallic counterparts carry higher risks of granulation tissue formation and fistulization. While these materials have not changed dramatically with the adoption of 3DP, the printing process enables improved anatomic congruence, especially for complex geometries or carinal involvement.

Polytetrafluoroethylene (PTFE), a flexible mesh material, has been incorporated at the Y-junction of intraluminal stents to enhance flexibility and accommodate the range of motion around the carina (n=4, 25.0%) (20,21,24,25). Another material used was polyurethane (n=1, 6.3%), selected for its soft, elastic properties and potential to minimize trauma to the airway mucosa (14). However, its clinical utility may be limited by hydrolytic degradation from respiratory secretions (29). Notably, PCL, a biodegradable polymer widely used in preclinical studies, was rarely employed in clinical cases (n=1, 6.3%). In one such instance, it was used to fabricate an external suspension scaffold rather than an internal circumferential implant (18). PCL is not currently approved by the United States Food and Drug Administration (FDA) for airway scaffolds in humans, and its clinical use requires individual case-by-case regulatory exemptions.

Clinical applications

Patients were categorized under one of the following etiologies: benign, malignant, or aerodigestive fistulas. Of the benign etiologies, post-lung transplant (LTx) anastomotic stenosis was the most frequent etiology. Post-LTx airway complications are not insignificant issues, with a reported prevalence of greater than 12% (30). The unique anatomy of LTx patients presents unique challenges when addressing these issues. Specifically, during LTx surgery, the length of the airway is often reduced to minimize ischemia (13). This practice often limits placement of standard commercial stents due to the proximity of the secondary carina and concerns of obstructing the distal airway (13,31). The first published report of a personalized 3D-printed stent airway was by Guibert et al., and it was successfully used in a patient post-LTx with a complete stricture of the bronchus intermedius and partial dehiscence of the bronchial anastomosis (17). In their study, the patient demonstrated an immediate improvement in symptoms and quality of life with no reported complications.

Since this first report, the use of 3DP stents for the treatment of post-LTx airway stenosis has grown with most reporting high congruence (12-15) (Figure 2). Overall, promising outcomes were observed as demonstrated by improvement in symptoms, quality of life, and pulmonary function tests. The second most common benign etiology was tracheobronchomalacia. Although tracheobronchomalacia results in significant collapse of the central airways, distortion of the native anatomy, resulting in significantly dilated airways (16,18) (Figure 3), the role of airway stenting in this condition remains debated and is generally reserved to stratify patients whom may respond to tracheobronchoplasty and for selected patients with severe symptoms who are poor surgical candidates. Conventional stents are frequently too small in diameter to provide an airtight fit, potentially necessitating the use of upsized self-expandable metal stents (32). However, metal stents cause inherent risks of metal fatigue, excessive granulation tissue, and fistulization, which increase with upsizing and increased pressures on airway walls (16,33). On the other hand, traditional silicone stents tend to exert less stress on tracheobronchial walls, although this may increase the risk of migration (33). Moreover, insertion of an oversized silicone stent can be technically challenging in patients with severe tracheobronchomalacia due to airway distortion in addition to possible airway injury in an already collapse distorted central airway. Given the challenge of balancing over- and under-sizing, there is a strong need for customizable airway stents in this population. Both external suspension scaffolds (18) and intraluminal stents (16) have demonstrated promise for the treatment of tracheobronchomalacia. However, a more recent study employing a silicone stent for tracheobronchomalacia resulted in suboptimal congruence, possibly due to poor capture of the baseline computed tomography (CT) scan if the patient was not in full forced inspiration (15). Notably, four-dimensional (4D) CT scans may be able to overcome this challenge by considering dynamic movements, thus providing a better mold for the eventual stent.

Figure 2.

Assessment and treatment of post-lung transplant airway stenosis with customized 3D-printed stents. (A) [1] Upper and [2] lower bronchoscopic views of personalized stent in a patient with post-transplant airway stenosis and [3] CT scan demonstrating congruence between the airway and the stent (denoted by white arrows). (B) [1] Pre-treatment 3D reconstruction highlighting stenosis (black arrow) and [2] post-treatment 3D reconstruction after virtual stent relief. [3] CT of the airways after stent insertion (white arrows) demonstrating congruence. (C) Pretreatment [1] CT and [2] 3D reconstruction showing previously migrated stent with granulation tissue as shown with black and white arrows in patient with post-surgical stenosis. [3 and 4] Design of the 3D-printed stent and [5] post-treatment imaging highlighting perfect congruence between the stent and airway as denoted by black arrows. (A-C) Reproduced from (15), with permission from BMJ Publishing Group Ltd. (D) Bronchoscopic view before, during, and after stent placement with 3D reconstruction of stent. (a) Bronchoscopic findings demonstrating airway stenosis. (b) Bronchoscopy after stent insertion. (c) Bronchoscopy after stent removal, showing airway remodeling. (d) 3D reconstruction of the airway prior to stent insertion, showing stenosis. (e) 3D reconstruction showing custom Y-stent diameters. (f) Custom airway stent prior to insertion. Reprinted from (13), with permission from Elsevier. 3D, three-dimensional; CT, computed tomography; Gr, granulation tissue; S, stent.

Figure 3.

Assessment and treatment of post-lung transplant tracheobronchomalacia using customized 3D-printed stents. (A) (a,b) Preoperative bronchoscopic and CT imaging compared to (c,d) bronchoscopy and imaging after treatment with tracheal suspension in a patient with tracheomalacia. (b) Reprinted from (18) under the terms of the Creative Commons Attribution—Noncommercial 4.0 License. https://creativecommons.org/licenses/by/4.0/. (B) (a) 3D modeling of defect for planning of virtual stent (shown in red) for patient with tracheobronchomalacia. (b) Dynamic reconstruction during forced expiration (left) and after implantation (right). (c) Comparison of a commercial stent (left) to 3D-printed stent (right). (d) Bronchoscopy after stent insertion showing device congruence with the native airway. Reprinted from (16), with permission from Elsevier. 3D, three-dimensional; Ca, carina; CT, computed tomography; LM, left main bronchus; RM, right main bronchus; Tr, trachea.

Management of malignant central airway obstruction (with or without involvement of the carina) is particularly challenging due to advanced tumor stage and poor overall health, preventing definitive surgical resection and instead necessitating palliative therapy. In these patients, intervention aims to relieve respiratory symptoms and improve their quality of life. Given current failure rates for all stent procedures (~22%) and the increasing prevalence of lung cancer globally, the need to improve airway stent technology is a pressing issue (34,35). Importantly, the role of customized 3D-printed stents in malignant obstruction is still in area of future investigation as standard stents are most of the times generally sufficient for cases. However, personalization may be beneficial in select situations involving complex carinal geometry or highly irregular tumor distortion and avoiding possible stent misplacement, secretion burden in an already frail population (25). In such cases, metallic, segmented Y-stents have been designed to fit at the junction of the main stem bronchus and its branches to overcome the dynamic challenges of the carina angle, providing an adequate match of the airway and significantly improving symptoms and respiratory status (21). This approach not only provides palliative relief but may also serve as a bridging therapy for those being considered for novel treatments. Beyond treatment, 3DP stents may be employed for stenosis prophylaxis in malignant settings (19).

Similarly, management of malignant laryngotracheal stenosis in inoperable patients can be a therapeutic challenge as it is a highly sensitive area subject to constant movement during phonation, coughing, or neck movements (36). Tumor invasion or the stent itself may trigger the cough reflex, increasing the risk of stent migration. Additionally, the laryngotracheal junction presents unique anatomical challenges as commercial stents may not perfectly match the transition between the larynx and upper trachea. This makes fine-tuning the size of the stent difficult, as under-sizing the stent reduces vocal cord stimulation and subglottic edema, while upsizing the stent prevents stent migration. Shan et al. demonstrated the feasibility of individualized 3DP barrel-shaped, metallic segmented transcordal stents and demonstrated a good match to the proximal airway with adequate patient tolerance, which resulted in improvement of symptoms and performance status, highlighting 3DP as innovative treatment for those patients who are not suitable for the current standard of care (20) (Figure 4).

Figure 4.

Metallic transcordal stent designed to fit both larynx and trachea in patient with laryngotracheal stenosis. Preoperative (a,b) and post-operative (c,d) CT (a,c) and 3D reconstruction (b,d). Reprinted from (20) under the terms of the Creative Commons Attribution—Noncommercial 4.0 License. https://creativecommons.org/licenses/by/4.0/. Schematic is created via BioRender with credit. 3D, three-dimensional; CT, computed tomography.

Aerodigestive fistulous tracts are another complex airway issue that has been the focus of recent clinical efforts. The use of airway stents in tracheoesophageal fistulas intends to isolate the airway from the esophagus, prevent further airway soilage, and control symptoms, either as a bridge for a future surgical reconstruction in patients who are candidates and/or transition them to definitive palliation in cases of malignant TEF (26,37). Furthermore, esophageal stents may be suboptimal given the fluctuations in diameter with contraction and relaxation (25). Several authors have demonstrated the feasibility of using indirectly printed metallic segmented Y-stents for the treatment of aerodigestive fistulas (n=3, 18.8%). These stents have shown success in achieving an adequate airway congruence, promoting fistula closure, and improving both symptoms and performance status in the majority of cases (24-26). In contrast, only one study (6.3%) to date has reported the use of a 3D-printed silicone stent for managing a tracheoesophageal fistula, in which the fistula failed to close possibly due to inadequate sealing from excessive bronchial wall pressures (23). These findings suggest that metallic stents may offer distinct advantages over silicone in this setting, particularly due to their ability to achieve more precise anatomical shape and their reduced risk of migration (25,38).

Toward regenerative solutions: insights from preclinical studies

Animal models

Most preclinical studies in the literature utilized rabbit models (n=41, 73.2%), given their ease of handling, cost-effectiveness, dimensions suitable for replacement, and similar histology to human tissues (39) (Table 3). Tracheal sizes do not vary significantly with weight in rabbits with endotracheal tube sizes similar to 3- to 9-month-old infants (65). Several studies utilized porcine models (n=11, 19.6%), which offer the advantage of more human-like airway dimensions and physiology. Sheep models were rarely utilized (n=2, 3.6%) despite the similarities to human anatomy and relative technical ease of implantation, possibly due to the compromised swallowing function in sheep, increasing risk of bacterial proliferation (58).

Table 3. Summary of preclinical studies and outcomes.

Article	Cell type	Total, n	Animal model	Time in vivo	3DP technique	Experimental scaffold material	Outcome
Park et al. (40)	Acellular	1	Pig	45 days	Direct	PCL with central core of polyethylene glycol diacrylate hydrogel containing erythropoietin	Vascularized tubular tissue flap with controlled luminal wall thickness by removal of a central core
							Heterotopic (latissimus dorsi) transplantation and erythropoietin containing hydrogel improved vascularization
Yu et al. (41)	Acellular	6	Rabbit	8 weeks	Direct	PCL frame with PCL nanofibers inside lumen and PCL microfibers outside of frame	Cell free biomimetic tracheal graft increased vascularization, smooth muscle infiltration and proliferation, and fibrous tissue deposition relative to conventional tracheal graft
Shan et al. (42)	Non-stem cells	Not stated	Rabbit	8 or 12 weeks	Direct	Hyaluronic acid methacryloyl-decellularized cartilaginous matrix composite hydrogel loaded with autologous chondrocytes in outer layer and PCL mesh coated with silk fibroin methacryloyl hydrogel loaded with autologous epithelial cells serves as inner layer	Epithelia-loaded silk fibroin methacryloyl hydrogel modified 3D-printed PCL mesh scaffold promoted regeneration of epithelium, cartilage, and vasculature
							Avoids need for preoperative (bioreactor) implantation
Ramaraju et al. (43)	Acellular	12	Pig	1 or 2 years	Direct	PCL scaffold, hydroxyapatite powder	PCL scaffold degradation after 2 years in vivo was 10% by mass and 50% by molecular weight
							Elastic modulus increased over 2× after 2 years in vivo and plastic strain was completely diminished by 2 years
Li et al. (44)	Stem cells	6	Rabbit	6 months	Direct	PCL scaffold with Matrigel seeded with FER-1 treated rabbit tracheal basal cells	FER-1 inhibits ferroptosis in tracheal basal cells, increasing epithelialization of scaffolds
Shai et al. (45)	Acellular	10	Pig	3 or 6 months	Direct	Silicone	Silicone tracheal grafts enabled growth of neocartilage
							Perichondral papillae, pre-resorptive layers, and vascular canals facilitated nutrient supply and waste removal
Shai et al. (46)	Acellular	32	Pig	Variable	Direct	Crystal graft with umbilical cord tissue, silicone, or PCL	PCL grafts exhibit improved integration and healing
							Addition of strap muscle reinforcement to the distal native trachea improved long term survival
McMillan et al. (47)	Stem cells	4	Ferret	5 weeks	Direct	PCL with GelMA with ferret MSCs	PCL scaffold with MSCs demonstrated healing of the defect site and epithelialization of the inner lumen
							Minimal chondrocytes were identified at the implant site
Krivitsky et al. (48)	Acellular	28	Rabbit	2 weeks	Direct	PLCL copolymer methacrylates alone with levofloxacin or nintedanib	Drug loaded stents locally administered medications while reducing systemic exposure
							Bacterial colonization was inhibited and local expression of IL-8 was reduced
Lee et al. (49)	Non-stem cells	6	Rabbit	4 or 8 weeks	Direct	PCL with chondrocyte-laden glycidyl methacrylated silk fibroin hydrogel	Pre-implantation into omentum for 2 weeks increased vascularization
							Scaffolds demonstrated effective tissue regeneration and luminal patency
Jin et al. (50)	Acellular	15	Rabbit	2 or 4 weeks	Direct	PCL and ethylene-vinyl chloride acetate copolymer with paclitaxel	Localized delivery of drug increased antitumor efficacy while reducing systemic toxicity
							C-shaped structure self-expanded and restored airway patency
Shai et al. (51)	Acellular	8	Pig	3 months	Direct	PCL	PCL scaffolds caused significant granulation tissue at the anastomosis requiring laser ablation to maintain patency
							Strap muscle reinforcement of the distal trachea stabilizes the implant and improves survival
Schleich et al. (52)	Acellular	21	Rabbit	8 weeks	Direct	PLCL methacrylates	A slightly conical, round shape with a helical structured outer surface facilitated transoral implantation and was well tolerated with minimal migration
Shan et al. (53)	Stem cells and non-stem cells	6	Rabbit	60 days	Direct	PCL and methacrylated silk fibroin methacryloyl hydrogel seeded with autologous tracheal epithelia	Increasing silk fibroin methacryloyl concentrations reduced scaffold degradation
						MSCs loaded on external surface after implantation	The scaffold underwent adequate epithelialization but did not undergo obvious chondrogenesis
Sun et al. (54)	Non-stem cells	4	Rabbit	8 weeks	Indirect, injection molding	GelMA with chondroitin sulfate methacryloyl seeded with chondrocytes and elastin methacryloyl as fibrous tissue-specific matrix gel surrounding a silicone tube seeded with fibroblasts	Tissue specific matrix hydrogels used to 3D print custom tracheal segments yielded biosimilar mechanical and biological environments
							Implanted tracheal segments had mature fibrous tissue with neovascularization
Tang et al. (55)	Non-stem cells	6	Rabbit	8 weeks	Direct	PCL with alternating rings of GelMA and chondrocytes	Implanted PCL trachea underwent transmural angiogenesis between cartilaginous rings
							8-week survival of 83% in transplanted rabbits
Liu et al. (56)	Acellular	20	Rabbit	1, 2, or 4 weeks	Direct	PCL with GelMA hydrogel with TGF-β and SDF-1	SDF-1 and TGF-β release promoted recruitment of local stem cells and differentiation
							Local cytokine releases reduced airway obstruction while increasing epithelialization
Wang et al. (57)	Acellular	15	Rabbit	30 days	Direct	PCL with nano-silica modification	3D-printed tracheal graft demonstrate superior biomechanical properties
							Decellularized tracheal grafts exhibit superior biocompatibility
Torsello et al. (58)	Stem cells	5	Sheep	6 months	Direct	PCL seeded with autologous MSCs	Excessive scaffold stiffness limits implantation and integration
							The cellularized PCL scaffold had poor integration in 3/5 cases
Pan et al. (59)	Stem cells	44	Rabbit	30 days	Direct	PCL with pluronic F-127 hydrogel carrying TGF-β seeded with BMSCs	Biologically active PCL scaffolds demonstrated improved cellular attachment and proliferation
							Bone marrow MSCs differentiated into chondrocytes and, mirroring native tissues
Kim et al. (60)	Acellular	2	Pig	3 weeks	Indirect, injection molding	Silicone	Right angle, triangular shaped outer rings increased friction between stent and airway
							In this limited study neither granulation tissue was not observed nor was mucostasis
Frejo et al. (61)	Non-stem cells	6	Rabbit	3 or 6 weeks	Direct	PCL with collagen and alginate with rabbit chondrocytes	Cartilage formation surrounding the scaffold was noted by 21 days with mature remodeling after 6 or 9 weeks
						PLA internal cylinder used for heterotopic transplantation	Explanted grafts demonstrated significant inflammatory responses which require further investigation
Pan et al. (62)	Stem cells	30	Rabbit	30 days	Direct	PCL seeded with MSCs	3D-printed biomimetic scaffold was immunogenic and demonstrated improved biomechanical properties
							While the 3D-printed scaffold had reduced biocompatibility relative to a decellularized graft, it had superior cellular adhesion and proliferation
Zhang et al. (63)	Stem cells and non-stem cells	20	Rabbit	6 months	Direct	Double layer PCL scaffolds seeded with chondrocyte outer layer and basal cell inner layer	Seeding of a 3D-printed scaffold with autologous tracheal basal cells enhanced epithelialization
							Epithelialized scaffolds improves early and mid-prognosis
She et al. (64)	Non-stem cells	6	Rabbit	8 weeks	Direct	PCL with a ring-hollow alternating structure and collagen sponge embedded in the hollows of the PCL rings	PCL scaffolds underwent remodeling by native chondrocytes which deposited tracheal cartilage, mimicking the structure and mechanical properties of the native trachea
Paunović et al. (39)	Acellular	6	Rabbit	2, 6, or 10 weeks	Direct	PLCL co-polymer methacrylates	After 10 weeks in vivo, stents demonstrated significant remodeling with normal morphology of ciliated columnar epithelium
							Round vs. slightly flattened design did not affect inflammation or tissue response
Wu et al. (65)	Acellular	30	Rabbit	4, 8, or 12 weeks	Direct	Mg-Li-Zn	Stents successfully increased tracheal lumen area
							Stents safely degraded in vivo ~8 weeks without affecting airway growth
Weber et al. (66)	Acellular	4	Pig	13–26 days survival	Direct	PCL with porcine-derived small intestine submucosa ECM patches lining the inside and outside of the graft	Flexible scaffolds portended ~10 days longer survival than rigid scaffolds but experienced increased granulation tissue deposition at the anastomosis
							Rigid scaffolds had reduced intraluminal granulation tissue
Weber et al. (67)	Stem cells	26	Pig	Variable	Direct	Bovine dermis ECM with MSCs, PCL with bovine dermis ECM, or PCL with porcine small intestine	Smaller circumferential coverage increased luminal epithelialization and survival
							Grafts embedded with stem cells supported chondrogenesis
							PCL groups developed significant intraluminal granulation tissue which can lead to obstruction
Lee et al. (68)	Acellular	18	Rabbit	24 weeks	Direct	PCL with Matrigel	2-week prevascularization in a platysma bioreactor increased luminal patency and long-term survival
							Matrigel underwent complete epithelialization by 12 weeks
Kim et al. (69)	Stem cells and non-stem cells	12	Rabbit	8 weeks	Direct	Silk fibroin methacryloyl with bilayer with tracheal basal stem cells in inner layer and chondrocytes in outer layer	4D silk fibroin methacryloyl hydrogel successfully mimicked tracheal tissues and adequately maintained airway patency
							Epithelialization was promoted along the inner lumen and chondrogenesis occurred at predefined locations that mimicked tracheal rings
Shai et al. (70)	Acellular and stem cells	10	Pig	5–19 days	Direct	VisiJet crystal plastic	There was a high mortality rate due to anastomotic occlusion or bleeding complications
							Chondrogenesis was present as early as 6 days after transplantation, although there was a concurrent substantial inflammatory response
Kim et al. (71)	Stem cells and non-stem cells	8	Rabbit	4 weeks	Direct	PCL with Matrigel seeded with iPSC-derived MSCs, iPSC-derived chondrocytes for outer layer and electrospun PCL seeded with bronchial epithelia cells for inner layer	Nanofiber tracheal graft embedded with induced pluripotent stem cell derived cells demonstrated similar biomechanical properties to the native trachea
							Epithelialization and chondrogenesis was promoted by induced pluripotent stem cell derived cells
Lee et al. (72)	Acellular	15	Rabbit	4 weeks	Direct	Polydopamine, polyethyleneimine, carboxylmethyl-β-cyclodextrin loaded with dexamethasone	Dexamethasone-loaded scaffolds reduced inflammation and enhanced tracheal mucosal regeneration
							Cartilaginous regeneration was not induced by the scaffolds
Chan et al. (73)	Acellular	10	Rabbit	0, 4, 5, 6, or 7 weeks	Direct	PCL	Implants underwent re-epithelialization near anastomoses but not at the center of the graft
							Significant granulation tissue led to airway obstruction and reduced survival
Hong et al. (74)	Non-stem cells	2	Rabbit	6 weeks	Direct	Silk fibroin with glycidyl-methacrylate hydrogel seeded with rabbit chondrocytes	Chondrocyte seeded hydrogel promoted epithelialization and chondrogenesis
							Biomechanical properties of the chondrocyte seeded hydrogel mimicked the native trachea
Townsend et al. (75)	Acellular	24	Rabbit	12 weeks	Direct	PCL and polylactide-co-caprolone with adhesion peptide or ceragenin-131	Adhesion peptide reduced luminal stenosis
							Polymer degradation rate affects tracheal patency and tissue overgrowth as demonstrated by the improved patency in the copolymer group
Gao et al. (76)	Non-stem cells	16	Rabbit	2 months	Direct	Poly(L-lactic acid) with reinforcing rings of poly(L-lactic acid) microfibers seeded with rabbit chondrocytes	In vitro pre-culturing with autologous chondrocytes and in vivo pre-vascularization in the sternohyoid effectively created a biosimilar scaffold
							Bipedicled muscle flap increased chondrocyte survival and accelerated epithelialization
Kang et al. (77)	Acellular	Not specified	Rabbit	2 or 4 weeks	Direct	Thermoplastic polyurethane wrapped in 4 layers of structural electrospun polylactide membranes, with one group having ionic liquid-functioned graphene oxide	Patterned electrospun membranes demonstrated improved mechanical properties, hydrophilia, and antimicrobial properties
							Patterned fibrous membranes promoted tissue formation
Xia et al. (78)	Non-stem cells	10	Goat	4 or 8 weeks	Direct	PCL C-shaped scaffold seeded with autologous chondrocytes	Autologous chondrocyte seeded PCL scaffolds prolonged survival when compared to an autologous tracheal graft
							While neovascularization and chondrogenesis did occur in the scaffolds, there was also stenosis and granulation tissue
Park et al. (79)	Non-stem cells	21	Rabbit	3, 6, or 12 months	Direct	PCL and alginate hydrogel seeded with epithelial cells on inner layer and chondrocytes in outer layer	Seeding with epithelial cells and chondrocytes prolonged survival
							Experimental scaffolds underwent re-epithelialization but did not promote cartilage regeneration
Park et al. (80)	Acellular	4	Rabbit	4 or 8 weeks	Direct	PCL	4-axis printing increased print accuracy and improved mechanical properties
							Scaffolds improved mucosal regeneration and reduced inflammation, but there was granulation tissue present
Kaye et al. (81)	Non-stem cells	6	Rabbit	3 or 6 weeks	Direct	PCL with hydrogel seeded with tracheal chondrocytes	Chondrocyte separation from the tracheal lumen with an intervening membrane promoted chondrogenesis
							Chondrocyte separation from the tracheal lumen with an intervening membrane reduced inflammation and stenosis
Pan et al. (82)	Acellular	17	Rabbit	15 or 30 days	Direct	PCL with nano-silicon dioxide surface modification	While porous PCL scaffolds initially underwent significant inflammatory reactions, this significantly abated after the acute phase
							Pore diameters of 200 microns promoted cellular adhesion
Park et al. (83)	Acellular	10	Rabbit	8 weeks	Direct	PCL scaffold cultured in omentum	Omentum cultured scaffolds demonstrated accelerated epithelialization and vascularization while reducing luminal stenosis
Park et al. (84)	Non-stem cells	20	Rabbit	2 months	Indirect, injection molding	PCL with silicone ring-shaped bands with tracheal mucosa decellularized ECM or collagen hydrogel on luminal surface seeded with human inferior turbinate mesenchymal stromal cell sheets	Complete luminal re-epithelialization can be achieved within 2 months
							Tracheal mucosa decellularized ECM directed differentiation of mesenchymal stromal cells into tracheal epithelium
Chen et al. (85)	Acellular	36	Rabbit	1, 4, or 8 weeks	Indirect	Silk fibroin	Scaffolds promoted epithelialization and vascularization narrowing and largely degraded by 8 weeks
							No significant granulation tissue ingrowth or luminal
Bae et al. (86)	Stem cells and non-stem cells	12	Rabbit	12 weeks	Direct	PCL with alginate hydrogel with inner layer seeded with epithelial cells and outer layer seeded with basal MSCs or chondrogenic-differentiated basal MSCs	Neo-cartilage formation was only noted when bone marrow derived MSCs were differentiated in chondrogenic media
							All PCL scaffold cohorts underwent neo-epithelialization and neo-vascularization in vivo
Townsend et al. (87)	Acellular	10	Sheep	10 weeks	Direct	Electrospun patch consisting of randomly layered PCL nanofibers enveloping PCL rings	Overgrowth of fibrous tissue resulted in tracheal stenosis
							Reactive changes including inflammatory and fibrotic changes were seen on the luminal side and surrounding the PCL support rings
Ghorbani et al. (88)	Stem cells and non-stem cells	9	Rabbit	4 weeks	Indirect, injection molding	PCL-collagen seeded with adipose-derived MSCs and primary chondrocytes surrounding decellularized aorta surrounding a PCL stent in the center	Collagen coated PCL and PCL blended with collagen demonstrated higher degradation than PCL alone
							Glycosaminoglycan release, type II collagen, and aggrecan expression were increased when PCL was blended with collagen
Gao et al. (89)	Non-stem cells	20	Rabbit	10 weeks	Direct	PCL seeded with chondrocytes	Scaffolds suspended in chondrocyte culture for 4 weeks prolonged survival when compared to those cultured for 2 weeks
							Both cohorts had a high mortality rate due to granulation formation within the scaffolds
Bhora et al. (90)	Acellular	5	Pig	17–34 days	Direct	PCL surrounding a circumferential ECM collagen layer	By 2 weeks in vivo, there was significant para-anastomotic granulation tissue with partial epithelialization
							270- and 360-degree scaffolds demonstrated equivalent efficacy
Rehmani et al. (91)	Acellular	7	Pig	3 months	Direct	PCL and dermal collagen ECM	Scaffolds with a dermal collagen ECM promoted respiratory mucosal proliferation and angiogenesis
							5/7 animals outlived initial 3-month study period
Jung et al. (92)	Acellular	32	Rabbit	4, 8, 12, or 16 weeks	Direct	Thermoplastic polyurethane with porous inner layer and non-porous outer layer	Scaffolds underwent re-epithelialization after 4 weeks with functional ciliated epithelium observed at 8 weeks
							Connective tissue ingrowth into scaffolds was observed after 4 weeks, with only one animal having mild luminal narrowing
Lee et al. (93)	Acellular	6	Rabbit	1, 4, or 8 weeks	Direct	PCL with Pluronic F127 membrane	Asymmetrically porous membrane design prolonged survival possibly due to permeation of nutrients and oxygen
							While anastomotic stenosis occurred in the asymmetrically porous membrane cohort, there was reduced luminal obstruction when compared to a bare PCL scaffold
Park et al. (94)	Stem cells	20	Rabbit	4 weeks	Indirect, injection molding	PCL and collagen gel seeded with tracheal mesenchymal stromal cell sheets	Human turbinate mesenchymal stromal cells promoted mature luminal epithelialization of the graft

3D, three-dimensional; 3DP, three-dimensional printing; 4D, four-dimensional; BMSC, bone marrow mesenchymal stem cell; ECM, extracellular matrix; FER-1, ferrostatin-1; GelMA, gelatin methacryloyl; IL-8, interleukin-8; iPSC, induced pluripotent stem cell; MSC, mesenchymal stem cell; PCL, polycaprolactone; PLA, polylactic acid; PLCL, poly(L-lactide-co-ε-caprolactone); SDF-1, stromal cell-derived factor-1; TGF-β, transforming growth factor-β.

Types of 3DP

In contrast to clinical studies, most preclinical grafts (n=50, 89.3%) were fabricated using direct 3DP methods that allow for layer-by-layer construction of scaffolds with precise control over geometry, porosity, and wall thickness, enabling high-resolution, defect-specific designs. In addition to design precision, direct 3DP may offer faster manufacturing times. Once imaging data is obtained, a printable model can be generated and rapidly fabricated (as fast as a few minutes to hours, depending on the size of the object, printer used, and process parameters), dramatically reducing lead times compared to conventional manufacturing or manual modification methods. Only a few studies employed indirect 3DP (n=6, 10.7%), in which molds were printed and used to cast other materials (54,60,84,85,88,94). This approach may be employed when the material of interest lacks the physico-chemical and/or rheological properties needed for direct 3DP.

Materials and design

While clinical studies primarily focused on non-biodegradable materials such as silicone or metal, preclinical studies in the literature have predominantly investigated biodegradable materials to facilitate regeneration of natural tissue growth while it subsequently and/or simultaneously degrades in vivo (n=54, 96.4%) (45,51,75). Silicone has favorable characteristics, such as chemical inertness, non-toxicity, and relative biocompatibility (45). However, its non-degradable nature limits its utility in regenerative models. One study utilized silicone bands or rings to reinforce scaffold architecture and mimic the native tracheal structure (84). However, standalone silicone scaffolds were uncommon (n=2, 3.6%), largely due to concerns regarding migration and lack of integration, which can lead to airway stenosis or obstruction over time (46). Similarly, metal-based scaffolds were used sparingly in preclinical research due to their lack of biodegradability and potential for tracheal injury (n=1, 1.8%). Wu et al. explored a magnesium-lithium-zinc alloy as a biodegradable metal scaffold (65). Their results showed promising outcomes: stents provided sufficient mechanical support, degraded completely within 8 weeks, and did not elicit systemic toxicity. Moreover, animals treated with this alloy demonstrated significantly larger airway lumens compared to those with the traditional, non-degradable metal stents. This could potentially help alleviate the need for device removal in traditional metallic stents. Nonetheless, biodegradable stents eliminate the risks associated with the presence of a chronic stent, as well as device removal (65). The ideal biodegradable material should exhibit robust mechanical properties that facilitate resistance to collapse after implantation, be biocompatible, available in unlimited quantities, and facilitate positive biomaterial-tissue interaction (92,95). To accomplish this, tissue-engineered trachea often employ a monolayer or multilayer scaffold frame composed of synthetic or natural materials and often incorporate additional components to promote cellular adhesion and growth (42,69,90).

PCL was the most utilized polymer for scaffold formation throughout the literature (n=42, 75%) (46,55,61,94). Its melting point makes it favorable for extrusion-based or fusion deposition modeling (91), while its slow degradation rate minimizes risks of collapse and restenosis associated with rapid degradation (43). These properties can be altered by adjusting the porosity, surface area, or through addition of other co-polymers (39,61,72). For example, poly(L-lactide-co-ε-caprolactone) (PLCL) combines faster degradation profile of polylactic acid (PLA) with the mechanical strength of PCL, allowing customization of scaffold longevity in vivo, and stiffness (75). Despite its advantages, PCL’s relative hydrophobicity presents a drawback, as it can limit cell adhesion and proliferation when used alone (73). To address this, surface modification techniques have been explored (57,72,82). Nano-silicon dioxide surface modifications have been shown to reduce hydrophobicity and enhance biocompatibility and cell attachment (82). Similarly, electrospinning PCL into nanofibrous structures may provide a more favorable environment for epithelialization and neovascularization (41,55,77). However, one study noted overgrowth of fibrous tissue and minimal tissue integration around electrospun patches, suggesting that incorporation of electrospun fibers alone may not be sufficient to support sufficient cellular adhesion and integration (87).

In addition to maintaining structural integrity and biocompatibility, 3DP scaffolds should also facilitate epithelialization and neocartilage formation to ultimately promote integration into host tissue. Unlike most tissues, cartilage lacks blood vessels and contains relatively few cells. Its unique viscoelastic behavior is primarily attributed to its extracellular matrix (ECM), which is composed mainly of water, collagen, and glycosaminoglycans (61). Addition of these materials to the scaffold has been shown to improve biocompatibility and reduce hydrophobicity in cases where PCL is being used (82). Hydrogels have since emerged as a versatile platform in tissue engineering, given their tunable mechanical stiffness, elasticity, biodegradability, and biocompatibility. These are polymeric materials with high water content that can form 3D networks with a highly porous structure, mimicking the natural ECM (42). They play a central role in tracheal tissue engineering by providing a hydrated, biocompatible matrix that supports cell adhesion, proliferation, and differentiation (68). A wide range of hydrogels are in use today, including those made from polysaccharides, proteins, synthetic polymers, and hybrid formulations.

Historically, due to the crucial role of collagen in cartilage structures, collagen-based scaffolds have been of interest for cartilage repair; however, they tend to degrade rapidly and lack mechanical strength, necessitating their combination with synthetic polymers or crosslinkers (90,91). Recently, methacryloyl hydrogels, which are photocrosslinkable polymers derived from natural proteins, have been the focus of many studies (42,47,49,54,55). One of which is gelatin methacryloyl (GelMA) (47,49,54,55) which is synthesized by functionalizing gelatin with methacrylate groups to promote cell adhesion. It has become a favored material due to its ease of use, excellent biocompatibility, and support for epithelial and chondrogenic differentiation (55). However, like collagen, it lacks mechanical robustness, which limits its standalone use in load-bearing applications. Similarly, silk fibroin methacrylate (SilMA), is a methacrylated form of silk fibroin that provides enhanced mechanical integrity (53). In addition to its use in hydrogel form, silk fibroin has been used in non-hydrogel scaffold forms, including electrospun fiber, films, or porous sponges (49,53,69). It is highly valued for its biocompatibility, minimal immunogenicity, and improved mechanical strength compared to GelMA (49,69,85). Moreover, silk fibroin has been shown to slow the degradation of PCL when used as a composite system (53).

Researchers have explored combining natural materials with more robust synthetic polymers like PCL to enhance both the structural and biological performance of scaffolds (49,53,59,88). For example, Ghorbani et al. compared three scaffold types: pure PCL, PCL coated with collagen, and PCL blended with collagen (88). Interestingly, they found that blending collagen into the PCL matrix significantly improved biocompatibility compared to surface coating alone, without compromising the strength of the scaffold. This may suggest that uniform distribution of natural materials throughout the scaffold may promote better cellular interaction and integration. More recent studies have employed this hybrid scaffold technique by co-printing GelMA with PCL (47). This multi-material approach allows for tuning of mechanical and biological properties within a single scaffold and is particularly valuable for constructs that must simultaneously support airway structure while encouraging tissue integration. Other strategies to promote cellular adhesion have involved the incorporation of bioactive signaling molecules, such as transforming growth factor-β (TGF-β) and stromal cell-derived factor-1 (SDF-1) (56,59). In a rabbit model, Liu et al. demonstrated that embedding TGF-β and SDF-1 into the scaffold significantly enhanced epithelial coverage and reduced airway obstruction compared to controls (56). Mechanistically, the presence of SDF-1 promoted the recruitment and chemotaxis of mesenchymal stem cells (MSCs) to the defect site, while TGF-β facilitated the migration of human bronchial epithelial cells (56). Notably, TGF-β was also found to promote the differentiation of MSCs into chondrocyte-like cells (56). The authors proposed that SDF-1 may have a synergistic effect with TGF-β in enhancing both cell migration and chondrogenic differentiation (56). These findings underscore the potential of functionalizing scaffolds with targeted growth factors to improve integration.

Cell seeding

Preclinical studies on 3DP tracheal grafts have utilized different approaches to create biocompatible scaffolds, generally falling into three main categories: acellular scaffolds, those seeded with non-stem cells, and those seeded with stem cells. Acellular scaffolds are typically fabricated from decellularized native tissue or synthetic polymers and are implanted without any cellular seeding. However, the absence of cellular components means there is limited capacity for active remodeling, signaling, or immunomodulation, which often leads to poor vascularization and delayed epithelialization (66,90,94). Despite incorporating biocompatible materials, these scaffolds often provoke an exaggerated immune response due to their inability to effectively communicate with the host environment (75,80,82,85). This immune activation can manifest as inflammation, fibrotic overgrowth, and luminal narrowing (46,66,79,93). Scaffolds seeded with differentiated, non-stem cells were the second most common approach in the literature and typically involved epithelial cells or chondrocytes. To mimic the complex architecture of tracheal tissue, a bilayer scaffold design can be employed where respiratory epithelial cells are seeded onto the luminal surface to recreate the mucociliary barrier, while chondrocytes are applied to the outer layer to support the cartilaginous structure (42,79). While these designs showed improved structural and functional outcomes compared to acellular scaffolds, they still faced challenges related to limited long-term cell viability and integration, as the cells lacked the regenerative capacity of stem cells (79).

Scaffolds seeded with stem cells generally show the most promising results for biocompatibility and tissue integration. These constructs often incorporate MSCs due to their multipotent differentiation capacity and immunomodulatory properties. When seeded onto 3D-printed scaffolds, MSCs can differentiate into both chondrocytes and epithelial-like cells under appropriate environmental cues. Studies involving these types of grafts generally showed successful reconstruction of tracheal defects with formation of mature and highly ciliated epithelium, neocartilage formation, and more robust neovascularization, which are key markers for functional airway restoration (47,59,62,84). A limitation of this approach is the unpredictable phenotype of stem cells, potentially resulting in insufficient cell amounts in vitro.

Several advanced models have explored the synergistic effects of combining both stem and non-stem cells in bilayered constructs to capitalize on the regenerative advantages of stem cells while preserving the specialized functions of terminally differentiated cell types (63,69,71) (Figure 5). Tracheal basal stem cells or epithelial cells are typically seeded on the inner lumen, while chondrocytes or MSCs are seeded on the external cartilage-supporting layer. Results from these models indicate enhanced graft stability, more consistent epithelial coverage, and improved structural integrity over time (63,69,71). The cell seeding process usually occurs via pretreatment of the scaffolds in vitro, followed by implantation. Given that longer pre-seeding durations increase efficacy (89), growth factors may be added to the cell culture medium in vitro to reduce seeding time and the risks of contamination and infection (69,86). Importantly, stem cells have an unstable phenotype, so there is a risk of obtaining insufficient numbers of the desired cell type. To address this, Zhang et al. created a double-layer scaffold that combined basal cells with exosomes in vitro to increase the speed of epithelialization in vivo (63). Ultimately, they found improved epithelialization, which subsequently decreased the incidence of late postoperative complications (63). One hypothesis is that exosomes improve cell proliferation via the inhibition of ferroptosis (44). However, the use of exosomes can be costly and complex, limiting suitability for larger applications. The same group aimed to address these challenges by culturing basal cells with ferrostatin-1 (FER-1), a ferroptosis inhibitor, and found that, similar to the use of exosomes, FER-1 decreased the amount of time needed for in vitro culture (44). Alternatively, in vivo heterotopic transplantation followed by orthotopic transplantation is a promising approach for improving cell seeding techniques (49). The heterotopic location first serves as an in vivo bioreactor (68), after which the implanted device can then be orthotopically transplanted as a tissue flap after removal of tissue in the core (40). Not only does this help promote cell maturation, but it has been shown to reduce stenosis at the implantation site and increase the rate of tissue regeneration (40,49,55,61).

Figure 5.

In vitro creation and in vivo intramuscular maturation of the 3D-printed tissue-engineered tracheal construct. (A,B) Chondrocytes and tracheal basal cells showed typical morphology and progressively covered the scaffold layers under dynamic and static culture, respectively. (C) Inner and outer layers were assembled, with the red dashed arrows indicating the movement of inner layer. (D) Constructs were implanted into the intramuscular pocket (black arrow) to allow for muscle encapsulation (green arrow). This construct may then be harvested as a pedicled muscle flap with vasculature (brown arrow). The native trachea is denoted by the white arrow. (E) Histology revealed cartilage formation in both groups, while only the experimental group developed an epithelial lining. Reprinted from (63), with permission from Elsevier. 3D, three-dimensional; H&E, hematoxylin and eosin; IF, immunofluorescence; SEM, scanning electron microscope.

Future directions and challenges

Looking forward, the development of 4D printing (4DP) may offer a way to bridge the gap between structural performance and biological responsiveness. These constructs are designed to change shape or function over time in response to environmental stimuli, such as temperature, pH, or osmotic pressure, allowing them to adapt their shape, stiffness, or surface properties over time (69,96,97). This adaptability may improve host integration by minimizing mechanical mismatch, enhancing dynamic interaction with surrounding tissue, and facilitating the release of bioactive molecules (98). Similarly, drug-eluting airway stents have the potential to deliver a wide range of agents, such as dexamethasone or antimicrobials, to directly address limitations of inflammation or infection (48,72), or to delivery drugs with high systemic toxicity, such as chemotherapeutic agents (50). Lastly, advances in computational modeling have further enabled the generation of a wide variety of sophisticated stent architectures through mathematical algorithms (99,100). Yilmaz et al. demonstrated that mechanical properties could be adjusted by changing mathematical constants in the design formula, without altering the material or printing method (99) (Figure 6). This level of control suggests that mechanical performance can be optimized entirely through geometry.

Figure 6.

Placement of a porous tracheal scaffold (A) in the trachea with various geometric configurations (B) designed using mathematical computation. Reprinted from (99) under the terms of the Creative Commons Attribution—Noncommercial 4.0 License. https://creativecommons.org/licenses/by/4.0/.

Beyond the laboratory, regulatory oversight remains a major barrier to clinical translation. Depending on their composition and function, 3D-printed products can be classified as medical devices, biologics, or combination products. Furthermore, medical devices are classified into three classes, with increasing regulation with higher class (101). For example, acellular scaffolds are likely to be considered class III medical devices (102). Unlike devices in the class II category, these devices are not eligible for the 510(k) pathway and require comprehensive data from clinical trials, often resulting in a longer and more expensive process (103). On the other hand, a scaffold seeded with cells or loaded with growth factors may be classified as a combination product, which is subject to much stricter oversight and different approval pathways compared to medical devices (104). Similarly, drug-eluting stents face substantial regulatory complexity because their embedded agents reclassify them as combination products rather than medical devices. This categorization dramatically increases the burden of proof for safety, efficacy, and manufacturing consistency.

Regardless of classification, FDA approval requires rigorous quality control to ensure that the manufacturing process produces consistent, reproducible, and safe products. This includes adherence to either the Quality System Regulation (QSR) for medical devices or Good Manufacturing Practices (GMP) for biologics and drugs. Combination products must often comply with overlapping regulatory standards, making the process even more challenging. Airway stents and scaffolds must demonstrate fatigue resistance and structural integrity over multiple life cycles with a high factor of safety. For example, although PCL is FDA-approved for certain applications, it is not approved for airway scaffolds in humans (105). Similarly, soft or non-load-bearing products, such as hydrogels, may be evaluated with more emphasis on degradation rate or viscosity. In addition, all implantable devices must undergo extensive biocompatibility testing for complications like cytotoxicity, irritation, and systemic toxicity. Drug-eluting stents and bioactive scaffolds must show controlled drug release or biological activity (106). While using well-characterized biomaterials can ease regulatory hurdles, many of the novel bioinks and resorbable polymers used in preclinical models lack established safety profiles. Additionally, the software used to design and produce 3DP devices is subject to regulatory scrutiny. Computerized design programs and data systems must all be validated to ensure reliability and cybersecurity (106). This is particularly important as the field aims to transition to direct 3DP.

Currently, most airway implants are used under custom device exemptions or compassionate-use provisions (102). Thus, there is a lack of large-scale safety and efficacy data, making it difficult to establish safety and efficacy benchmarks. Without standardized clinical protocols and long-term outcome tracking, the field struggles to meet the burden of proof required for regulatory approval. Altogether, while 3DP holds significant promise for personalized and regenerative medicine, the regulatory landscape remains a formidable barrier to clinical translation.

This review is limited by several factors, including the number of available clinical studies, with most reports involving single-patient trials or case series, which limits generalizability. The majority of the preclinical and clinical studies lack long-term follow-up and standardized outcome measures, making it difficult to assess stent durability and safety over time. Because our search strategy was limited to PubMed, relevant engineering-focused studies indexed in other databases may not have been captured. In addition, because this work is not a systematic review, a formal quality assessment of individual studies was not performed, and variability in methodological rigor among the included literature should be considered when interpreting the findings. Ultimately, successful translation of 3D-printed airway constructs will depend on sustained multidisciplinary collaboration among engineers, scientists, clinicians, and regulatory bodies. Designing early-phase clinical trials with long-term outcome monitoring and publishing standardized protocols will be key to advancing this technology within the clinical space.

Conclusions

This review underscores the transformative potential of 3DP in advancing personalized airway reconstruction. For clinicians, 3DP offers the ability to design patient-specific implants that achieve superior anatomical congruence and improved functional outcomes compared with conventional devices, particularly for patients with complex airway disease where standard stents and surgical options are limited. From a policy and implementation perspective, the integration of 3DP into clinical workflows will require clear regulatory pathways and standardized manufacturing protocols to support wider clinical adoption beyond single-patient compassionate use. Health systems and regulatory bodies must work collaboratively to balance innovation with safety and quality assurance. For researchers, the field demands long-term outcome studies, rigorous comparative trials, and continued translational efforts to bridge preclinical advances in bioactive, biodegradable scaffolds with real-world clinical application. Further work on optimizing manufacturing speed, ensuring reproducibility, and developing scalable tissue-engineering strategies will be essential. Together, these efforts can facilitate a structured pathway toward making personalized 3DP airway reconstruction an accessible treatment approach for complex airway disease.

Supplementary

The article’s supplementary files as

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.