Tissue Engineering in the Trachea


  • Koji Kojima,

    Corresponding author
    1. Laboratory for Tissue Engineering and Regenerative Medicine Department of Anesthesiology, Harvard Medical School, Boston, Massachusetts
    • Correspondence to: Koji Kojima, Laboratory for Tissue Engineering and Regenerative Medicine, Department of Anesthesiology, Harvard Medical School, Brigham and Women's Hospital, 75 Francis Street, Thorn 703, Boston, MA 02115. E-mail: kojima@zeus.bwh.harvard.edu

    Search for more papers by this author
  • Charles A. Vacanti

    1. Laboratory for Tissue Engineering and Regenerative Medicine Department of Anesthesiology, Harvard Medical School, Boston, Massachusetts
    Search for more papers by this author


This article is corrected by:

  1. Errata: Corrigendum: Tissue Engineering in the Trachea Volume 299, Issue 6, 703–710, Article first published online: 18 April 2016


This review summarizes efforts to generate an autologous tissue-engineered trachea (TET) using various biomaterial or cell sources to make tracheal cartilage to form the structural components of a functional tracheal replacement. Biomechanical assessments of the TET showed that the cartilage stiffness was excellent in the nude models; however, the sheep autologous TET did not provide sufficient support and collapsed easily. As a result, tissue engineering technology is still far from allowing the functional recovery of patients who suffer from severe tracheal disease. On the other hand, there are several clinical reports seeding cells to decellularized tissue using tissue engineering techniques. However, the working mechanisms of the tissue-engineered trachea remain unclear. Nevertheless, we believe that the field of tissue engineering has great potential for surmounting these obstacles and allowing us to generate functional tracheal replacements in the near future. Anat Rec, 297:44–50. 2014. © 2013 Wiley Periodicals, Inc.


This review provides a summary of the progress made over the past two decades in the generation of a tissue-engineered tracheal equivalent. A plethora of efforts have been attempted over the last 20 years to generate tracheal equivalents (Grillo, 2002). Direct anastomosis currently serves as the gold standard therapy for patient with tracheal disease, even though it is known to have several limitations. For instance, in adults, only half the tracheal length can be successfully treated by tracheal resection and anastomosis, while in pediatric patients the limit drops to about one third. This issue becomes challenging for physicians because several ailments, including tracheal stenosis secondary to post-intubation tracheal granulation, tumors, or cancers, lead to defects greater than those amenable for direct resection and anastomosis repair. Presently, the patients that fall outside the aforementioned therapy limits are treated with suboptimal laser treatments and stenting. Additionally, experimental treatments, including the shaping of a tracheal conduit from a variety of tissues and materials, such as autografts (Letang et al., 1990; Cavadas et al., 1998; Osada and Kojima, 2000), allografts (Messineo et al., 1994; Kushibe et al., 2000), prosthetic materials (Graziano et al., 1967; Okumura et al., 1991) or a combination (Osada et al., 1994; Banis et al., 1996), have also been attempted. Nevertheless, the use of these techniques also has restrictions. Currently available prosthetics frequently have problems, such as the inability to mimic the properties of native tissues, as well as suffering from extrusion, incomplete host-tissue integration and inflammation. A tracheal transplant is another option, but donor tissue is difficult to obtain.

As of today, these techniques have not proven to be practical. Thus, the use of tissue-engineering techniques utilizing autologous stem cells to create new tissues is ideal (Vacanti et al., 2001). Tissue engineering merges biology, engineering, and materials science to generate new tissues in vitro for later use in vivo (Bonassar and Vacanti, 1998) so as to allow in vivo growth of in vitro-produced tissues. Therefore, tissue engineering has provided an additional way of generating replacement cartilage structures.


For the generation of a viable engineered trachea, the process must use relatively small initial tissue sources, from which cells can be significantly expanded to the numbers necessary for tracheal tissue generation. Cells should be maintained in culture that promotes growth and replication, as well as preventing dedifferentiation from the chondrocyte phenotype. Upon achieving adequate cell numbers, scaffold selection is important, as it will serve as the shaping material and delivery vehicle of the cells to the recipient site. Materials available include biocompatible and biodegradable polymers, each with its particular advantages and disadvantages. Regarding tracheal tissue engineering, biomaterials should be adequately biodegradable and hydrogels should be easily absorbable over time, metabolized in the liver and/or excreted in the urine. Biocompatible, non-biodegradable materials were not discussed in this review, such as silicone, ceramics (hydroxyapatite), metals (titanium), and composite hybrids. We instead discuss bio-artificial tracheal prosthetics. Finally, upon implantation, seeded cells must survive and express the appropriate genetic markers for cartilage, even in immunocompetent hosts. Only in these conditions can the newly generated tracheal tissue be tested for strength, shape, and other mechanical properties required for clinical use in human patients.

The first attempt forcusing on tracheal cartilage using tissue engineering techniques was reported by Vacanti et al. (1994). Soon after, many studies started to report on the reconstruction of the trachea (Fuchs et al., 2002; Kojima et al., 2002; Kojima et al., 2003b; Zang et al., 2012; Zani et al., 2008). Lee et al. (2002) showed that a 5 × 5-mm full-thickness excision of the rabbit tracheal wall could be substituted with a cell-seeded implant scaffold. The authors observed epithelialization, but were incapable of detecting viable chondrocytes from the implants. Various other groups have also reported limitations from tissue-engineered tracheal replacement techniques. Fuchs and group showed that prenatal tracheoplasty utilizing engineered cartilage can be an effective treatment for severe congenital tracheal malformations, yet it showed limited use in adults (Fuchs et al., 2002).

Our group believes that the tracheal cartilage component is key to its structural support. The trachea is somewhat unusual in that its cartilage portion only encompasses about two-thirds of the circumference and the remaining one-third is made of smooth muscle. Moreover, the flexibility and stiffness of the cartilage in the trachea is a key factor in elucidating the cross-sectional area of the airway during forced expiration dynamic collapse. Additionally, the trachea is an organ continuously exposed to the outside environment, leading to increased risk of infection, vascular erosion, and extrusion. Previous attempts at in the development of tracheal replacements have found the aforementioned obstacles hard to conquer (Osada et al., 1994). However, by developing a functional tissue-engineered trachea, surgeons will be able to prevent patient demise and improve outcomes, something current treatments cannot always achieve. The use of tissue culture and engineering concepts, along with the adequate understanding of scaffold properties and preparation, can provide a solution to problems in the field.


Harvesting of cells for the generation of functional cartilage structures is a major challenge in tracheal reconstruction. Ultimately, the goal is to generate a large number of cells with a universal phenotype that would be non-immunogenic and able to be used with any patient. Currently, the most effective source of cartilage is still unknown. Preferably, this source should require minimally invasive harvest procedures that would still yield sufficient cells to allow the creation of a tissue-engineered trachea. Currently, autologous cell use is ideal because it obviates the need for immunosuppression of recipients to block tissue rejection. The current understanding of stem cell biology, including mesenchymal stem cell (MSC), embryonic stem (ES) cells, and induced pluripotent stem (iPS) is very encouraging for the field of regenerative medicine. In addition, the potential ethical issues arising from the use of these cell types must be considered. Currently, there is no consensus regarding the ideal cell type for tracheal tissue engineering. The current focus is on the use of autologous chondrocytes or mesenchymal stem cells for regenerating trachea.

One of the most important issues in tissue engineering cartilage is devising the optimal harvesting technique that affords simplicity, safety, and minimal invasiveness. Furthermore, the source of cartilage must also be optimized. Currently, the trachea, nasal septum, knee, and ribs are the anatomical sites where hyaline cartilage has been obtained for the study of cartilage tissue engineering. Ear cartilage, composed of elastic cartilage, is also a possibility. It is the simplest site from which to harvest tissue for engineering purposes. The ear fulfills the necessary criteria for a cartilage source. However, the use of elastic cartilage cells for the engineering of hyaline, articular cartilage has been debated. On the other hand, nasal septum cartilage has similar properties to that found in tracheal cartilage. Chondrocytes, epithelial cells, and connective tissue can all be obtained from the same small biopsy of nasal septum, as our group has previously shown (Kojima et al., 2003a). Nevertheless, additional sources must be explored. Recently, adult stem cells have been shown to have huge potential, and tissue engineers have renewed their interest on these cells. Our group believes that MSCs have the greatest potential for the development of therapies within cartilage tissue engineering. We have recently studied the possibility of using MSCs as an alternative to nasal chondrocytes. MSCs can be harvested with techniques even less invasive than nasal septum biopsies, such as fine-needle aspirations. Our laboratory is currently studying the differentiation ability of MSCs toward cartilage for tracheal engineering purposes. Because of the huge potential of bone marrow-derived mesenchymal stem cells derived for regeneration of tissues, their properties are being studied within the field of bioengineering, especially in the orthopedic field for cartilage, bone, tendon, and ligament regeneration.

However, the use of MSCs alone is still limited in the clinic due to the need of additional growth factors for differentiation, such as TGF-b1–3, dexamethasone, or ITS supplement. In order to avoid additional approval obstacles from the FDA or an ethics committee, we explored the feasibility of using primary cells and MSC co cultures to trigger differentiation and repair. We demonstrated that anterior cruciate ligaments cells (ACLs) and MSC co-cultures at a 1:1 ratio exhibited enhanced ligament repair compared to either population cultured independently (Canseco et al., 2012).

Based on our ligament results, we preliminarily explored the co-culture of MSCs and chondrocytes in a 1:1 ratio, finding that cartilage quality was improved in co-cultures compared to the use of nasal septum cartilage alone (data not shown).


Once the donor cells have been harvested, expanded, and the chondrocytic phenotype maintained, the next step is to deliver them to the recipient host. There are currently many potential scaffolds for use in tissue engineering that the challenge of choosing the correct material is enormous. Large categories include biodegradable polymers (e.g., polyglycolic acid, polylactic acid, and polycaprolactone), hydrogels (e.g., Pluronic F-127, collagen gel), and decellularized matrix scaffolds. In terms of desirable properties of these materials, of course, it is preferable for the material to not cause an inflammatory reaction either in its natural state or in the form of breakdown products, and it should not have toxic effects on the donor or recipient cells. Furthermore, it should be a good substrate for cell attachment, so that cells are not lost from the desired implantation site. Previous studies have shown that cells injected directly into the host without a scaffold will disperse away from the injection site, such that the concentration is insufficient for subsequent tissue formation (Cao et al., 1998). The material should also have a porosity that allows cells and nutrients to intercalate evenly throughout the material. Without sufficient pore size, distribution or interconnection, one risks creating scaffold-shaped shell of tissue, due to cell death caused by malnutrition or lack of cell dispersion to the center of the material. Pore size has also been shown to affect the differentiation of the attached cells, with studies on osteogenesis showing smaller pore sizes lead to osteochondral formation and larger pore sizes result in direct osteogenesis (Karageorgiou and Kaplan, 2005). Finally, the scaffold should ideally degrade at a rate that is matched with neocartilage development.

Our laboratory and others have compared nasal chondrocytes seeded in hydrogel, PGA and polycaprolactone in immunodeficient rodent models. In particular, our results showed that PGA with chondrocytes had better results in vivo and in vitro (unpublished data). As a result, we believe that PGA is a good material for tracheal tissue engineering. In addition, PGA is already an FDA approved suture material, which enhances its attractiveness for clinical applications.


We studied both sheep tracheal and nasal septum chondrocytes to determine the harvesting cell yield and the quality of the engineered cartilage. For the study, 5 × 5-mm samples of sheep nasal septum cartilage (N = 6) and two to three tracheal rings (N = 6) were obtained from 2-month-old sheep. Both cell–polymer constructs were then incubated in vitro for 1 week and then wrapped around a silicone tube of 7 mm diameter × 30 mm length, which was subsequently implanted into a subcutaneous pocket in a nude mice. The implants were recovered at 8 weeks and analyzed histologically, biochemically, and biomechanically. The tissue was assayed for cartilage-specific extracellular matrix components, including proteoglycans and collagen. In this study, the cell-polymer constructs formed cartilage de novo in vivo in the shape of cylinders after 8 weeks. The gross appearance of both the tracheal-chondrocyte-derived cartilage and the nasal septum-chondrocyte-derived cartilage tissue-engineered tracheas (tracheal TET, nasal TET) were very similar to that of native tracheal cartilage. Each exhibited a translucent white appearance reminiscent of the hyaline cartilage of native trachea. The gross consistency and elasticity were also comparable. Both samples had a cartilaginous histology similar to that of native tracheal cartilage. In addition, the proteoglycan content of the tracheal TET, nasal TET, and native tracheal cartilage were 84.3 ± 7.5 µg/mg, 97.1 ± 3.2 µg/mg, and 120.0 ± 9.2 µg/mg, respectively. The collagen content of the tracheal TET, nasal TET, and native tracheal cartilage were 1.25 ± 0.21 µg/mg, 1.28 ± 0.20 µg/mg, and 1.36 ± 0.13 µg/mg, respectively. The tensile modulus of native trachea was 10.6 ± 1.8 MPa, significantly higher than that of tracheal TET (1.4 ± 0.4 MPa) or nasal TET (1.4 ± 0.5 MPa). There was not a significant difference between the tensile moduli of tracheal TET and nasal TET. Thus, we concluded that nasal septum cartilage is a potential site for obtaining an adequate sample of tissue for further tissue-engineering studies (Kojima et al., 2003a). The advantage of using hyaline cartilage from the nasal septum is that it not only has properties similar to those of tracheal cartilage, but is also a source of epithelial cells and connective tissues (Kojima et al., 2002). Nasal epithelial tissues were separated from the underlying nasal septum cartilage. Epithelial cells were also obtained from the mucosal lining of the nasal septum. Nasal epithelial cells at a concentration of 50 × 106 cells/mL were suspended in 23% Pluronic F-127 and injected into the implanted cylindrical tube of cartilage that was generated in the nude mouse around the silicone tube template. Cell–polymer constructs formed de novo cartilage in the shape of cylinders lined with a pseudo-stratified columnar epithelium after 10 weeks (Fig. 1). The obvious benefit is that one is able to obtain from the same small nasal septum biopsy the three cell types necessary for in vitro three-dimensional tissue construction. However, cartilage harvested from nasal tissue should not be considered the universal or unique source for all cases of tracheal injury in which tissue engineering could be beneficial. For instance, the availability of nasal cartilage may be limited in children or victims of smoke inhalation, and its use as a source of cartilage could be counterproductive. For those non-ideal patient cases, there may be an advantage in using MSC, as obtaining them is less invasive than removing a sample of nasal cartilage. We evaluated the feasibility of using autologous sheep mesenchymal stem cells cultured onto a PGA mesh to engineer a helix-shaped cartilage equivalent of a functional trachea (Kojima et al., 2004). Transforming growth factor-β (TGF-β) has been shown to play a major role in cartilage development, and studies have demonstrated that TGF-β2 helps support chondrogenesis in developing 3D tissue constructs in vitro. In vivo, we explored the potential benefit of local delivery of TGF-β2 to cells using biodegradable microspheres. We were able to deliver TGF-β2 locally by coupling its release to the degradation of a biodegradable hydrogel prepared by cross-linking acidic gelatin with glutaraldehyde. This gelatin hydrogel incorporating TGF-β2 effectively promoted cartilage regeneration in vivo.

Figure 1.

Tissue (a) engineered trachea appearance, (b) engineered trachea cross-section, (c) nasal engineered trachea, and (d) nasal engineered trachea cross-section.

Sheep bone marrow was obtained by iliac crest aspiration from 6-month-old sheep. Cell suspensions were concentrated to 50 × 106 cells/mL and seeded on 100-mm × 10-mm × 2-mm non-woven meshes of PGA fibers. The cell-seeded mesh samples were cultured continuously for 10-week in vitro. For the in vivo study, the cell-seeded mesh samples were cultured continuously for 1-week in vitro. The structures were then placed in the grooves of a helical template 20 mm in diameter × 50 mm in length made with a Silastic ERTV mold-making kit. Cell–polymer samples were coated with gelatin microspheres with TGF-β2. The composite structures were implanted subcutaneously in immunodeficient rats and harvested after 6 weeks. Both engineered cartilage samples were solid and shiny white with a formed cartilaginous circular helix. The TETs showed great similarity to native trachea and, like native cartilage, were stiff to the touch, yet flexible (Fig. 2).

Figure 2.

Tissue-engineered trachea (TET) cross-sectional appearance (a) with cartilage and pseudoepithelial lining. Hematoxylin and eosin stain showed the mature cartilage and the formation of a layer of psuedoepithelium (b).


Subsequently, we focused on the creation of an autologous tissue-engineered cartilage construct shaped as a helix to form the structural component of a functional tracheal replacement, with tracheal epithelial cell sheets. 5 × 5 mm samples of sheep nasal septum were obtained from 2-month-old sheep as previously described (Kojima et al., 2003a). Chondrocytes were isolated by digestion of cartilage in collagenase, and epithelial cells were obtained from the mucosal lining of the same sample. After 2 weeks in culture, epithelial cells were stored and chondrocyte suspensions were placed on 100 mm × 10 mm × 2 mm PGA mesh fibers. This chondrocyte-seeded mesh was placed in the grooves of a 20 mm diameter × 50 mm long helical silicon template (Fig. 3a) and implanted under the sternocleidomastoid (SCM) muscle of the corresponding sheep (Fig. 3b). Eight weeks post-implantation, the silicon template was removed from the autologous TET while keeping the vascularized TET connected to the SCM muscle pedicle (Fig. 3c). Then, a 7 cm circumferential cervical trachea segment was excised, and the autologous TET was transplanted to the site by an end-to-end anastomosis. The engineered epithelial cell sheet (2 weeks prior to transplantation, epithelial cells were cultured on temperature-responsive culture inserts) was wrapped around the external surface of the TET, including the site of anastomosis (Fig. 3d). A silicone stent (10 cm) was inserted before completing the distal anastomosis. Internal coverage of the entire length of the TET by the stent was confirmed by bronchoscopy, and the stent was secured in place to prevent migration with two sutures placed at its oral end. Prior to closing the surgical site, the entire vascularized construct was covered with the SCM muscle pedicle. Sheep were euthanized at 4 weeks and the TET was evaluated. The sheep tolerated the surgical procedure well with no perioperative complications. The proximal and distal anastomosis sites of the TET transplant were clearly visible through the transparent stent, and exhibited no problems at 1 week by bronchoscopy. The gross morphology of the TET was a white, shiny, hard tissue with confirmed epithelialization along its entire length, including both anastomosis sites (Fig. 4). However, cartilage content of the tissue was less than 10% of the TET.

Figure 3.

Appearance of TET in vitro for 10 weeks (a), lateral view (b), and cross-section (c).

Figure 4.

Implantation of a tissue-engineered trachea in a sheep model. A helical template was made utilizing a silicone mold kit. (a) The chondrocyte-seeded matrix was positioned in the grooves of the silicone template. (b) The implant was sheltered by the sternocleidomastoid muscle. (c) The tissue-engineered trachea was harvested from the neck of the sheep 8 weeks after the procedure with muscle pedicle. The lumen of engineered trachea showed epithelium on the engineered surface.

We have demonstrated that a long, circumferential tracheal defect can be successfully transplanted with a TET covered with an epithelial sheet and supported by an inert stent. The external wrapping of the epithelial cell sheet is a key point in our study (Zani et al., 2008)


The trachea is complex organ because of its particular mechanical properties and specialized function. Our laboratory group believes that in order to successfully develop a tissue-engineered trachea the following requirements should be met: (1) develop a minimally invasive and simple harvest procedure that yields sufficient cells in a short amount of time; (2) the use of a biodegradable scaffold that does not cause inflammatory and immune responses; (3) proper flexibility, stiffness, and compliance of the engineered tissue; (4) an airtight structure; (5) appropriate angiogenesis and vascularization of the tissue; (6) epithelialization of the tissue; and (7) growth of the implant in the patient, particularly important in the pediatric population.

Our group believes that living cells must be used for the successful generation of tissue-engineered tracheas; thus a lot still needs to be elucidated regarding cell morphogenesis and biomaterial polymer scaffolds, particularly their biodegradation properties. Furthermore, it is imperative to learn about cell–biomaterial and tissue–biomaterial interactions and their behavior in the setting of wound-healing responses. As we further our knowledge of the normal development of tissues, the interactions between tissue and biomaterials, and the wound-healing response, we may be able to take advantage of native tissue properties for the development of engineered living tracheal tissue. A foremost challenge in the field of tracheal tissue engineering is the harvesting of living cells. The main goal in this regard is the elucidation of a method to treat critical tracheal defects with tissue-engineered implants developed using large-scale techniques. In an ideal future, large numbers of cells would be obtained from a universal source from a patient's own body, avoiding any negative immunologic responses. These cells could come from multiple sources, including autologous cartilage, adult stem cells, ES cells, iPS cells, and MSCs, all of which may be adequate for the development of engineered tissues.

There have been four published human cases over the past 4 years (Macchiarini et al., 2008; Delaere et al., 2010; Jungebluth et al., 2011; Elliott et al., 2012). Jungerluth et al. (2011) reported the use of a bioartificial nanocomposite with stem cells in a 36-year-old male patient for tracheobronchial transplantation in a case of mucoepidermoid carcinoma. Even though this is not a true tissue engineering approach, because of the use of an artificial prosthesis, the investigators did use bone marrow stromal cells, which is a novel incorporation to their technique. However, it should be noted that in a patient with carcinoma, long-term follow-up is important, in case dormant cancer stem cells exist in their bone marrow. Two additional studies, by Macchiarini et al. (2008) and Elliott et al. (2012) used allogeneic and enzymatically decellularized tracheas for replacement in a 30-year-old female and 10-year-old child, respectively. Macchiarini et al. used autologous epithelial cells and chondrocytes, while Elliott et al. used bone marrow derived mononuclear cells. Both reports were unclear on the mechanism by which the cells repopulate decellularized trachea. Regarding the decellularized trachea, the decellularization process is crucial, especially if cells or DNA residue remains in the tissue, which can cause immune and inflammatory responses. In addition, decellularization has been observed to negatively affect the ECM component of the tissue, including the main structural elements of cartilage, including proteoglycans and collagen. These effects can lead to poor structural stability in the tracheal replacement, as shown by Elliott et al. In their study, they initially used a biodegradable stent, which failed after a period of 5 months, and had to be replaced by a permanent nitinol stent to stabilize the transplant. In addition, using a decellularized allogeneic graft as scaffold material will not address the problem of donor organ shortage. Although decellularized technology is rapidly progressing in the regenerative medicine field, its optimal implementation might rely on the usage of xenogeneic decellularized tissues, once we can reliably eliminate DNA and residual cell contamination.

At this point, none of the tissue-engineered trachea studies have been clinically successful. The most recent procedure for a patient needing long tracheal defect replacement used allotransplantation. Delaere et al. reported that tracheal allotransplantation was performed with a vascularized allograft after indirect revascularization of the graft was performed in a heterotopic position with concomitant immunosuppressive therapy. It was a very long procedure that required immunosuppressive therapy before transplantation, but is the first choice therapy for patients needing long tracheal replacement. A recent report by Luo et al. (2013) showed positive long-term function of engineered tracheas with pedicle vascularization in a rabbit model. So, it is exciting to see which procedure is more suitable for patients as new technologies are developed with cell culture methods and novel materials.

In conclusion, in order to be able to transfer our tissue engineering techniques into the clinic for the benefit of patients, it is also very important to develop appropriate surgical procedures, and ideal cell culture conditions and preparation. Furthermore, the use of alternative biomaterials as scaffolds for cartilage growth, and the procurement of vascular supply to the implant needs to be examined as well.