Successful immunosuppressant-free heterotopic transplantation of tracheal allografts in the pig

Abstract

OBJECTIVES: It has been demonstrated that both heterotopic and orthotopic transplants of epithelium-denuded cryopreserved tracheal allografts are feasible in immunosuppressant-free rabbits. Validation of these results in large animals is required before considering clinical applications. We evaluated the viability, immune tolerance and strain properties of such tracheal allografts heterotopically transplanted in a pig model.

METHODS: Ten tracheal segments, 5 short (5 rings) and 5 long (10 rings), were obtained from male Landrace pigs. The tracheal segments were surgically denuded of their epithelium, then cryopreserved and stored in a tissue bank for 33 to 232 days. After thawing, tracheal segments stented with a silicone tube were wrapped in the omentum in 2 groups of 5 female recipients. The animals did not receive any immunosuppressive drugs. The animals were euthanized from Day 6 to Day 90 in both groups.

RESULTS: An effective revascularization of allografts regardless of length was observed. Lymphocyte infiltrate was shown in the early postoperative period and became non-significant after 30 days. Allografts displayed high levels of neoangiogenesis and viable cartilage rings with islets of calcification. Biomechanical measurements demonstrated strain properties similar to those of a fresh tracheal segment from Day 58.

CONCLUSIONS: Our results demonstrate the acceptability and satisfactory stiffness of epithelium-denuded cryopreserved tracheal allografts implanted in the omentum, despite the absence of immunosuppressive drugs. Since the omentum has the capability to reach the tracheal region, this approach should be investigated in the setting of orthotopic transplants in a pig model before considering clinical applications.

INTRODUCTION

Despite clinical successes, replacing more than half of the trachea remains a formidable challenge [1, 2]. Finding the ideal substitute for a fully circumferential tracheal replacement is ongoing. A tracheal allograft appears to be a possible option since the first short-term follow-up report in 1979 [3]. Since then, only 2 fully circumferential orthotopic tracheal transplants (OTTs) have been reported in patients [4, 5]. Indeed, OTTs have been regarded as a failure due to stenosis, graft malacia or necrosis imputed to varying degrees of immune rejection and ischaemia [6, 7]. Failure was also compounded by the characteristics specific to the trachea: biomechanical properties, exposure to the outside world and absence of well-defined blood supply impeding an OTT with immediate revascularization as performed during transplant of solid organs. To resolve this issue, indirect heterotopic revascularization of tracheal allografts using a fascial flap wrap has been investigated in a rabbit model, allowing successful 2-stage OTT [8]. However, the clinical application of this staged procedure with the graft revascularized first using a forearm fascial flap pedicled on the radial artery and veins led to tracheal membrane necrosis, thus limiting the free flap procedure to non-circumferential OTT [2].

Otherwise, the feasibility of an omentum flap-wrapped autograft transplant has been assessed in canine models. These studies showed that the success or failure of the procedure was dependent on graft length [9–11].

Another major issue is the immunogenicity of tracheal allografts, which requires immunosuppressive drugs clearly contraindicated in cancer patients, thus restricting OTT to benign extended tracheal lesions [3–5]. Given that the respiratory epithelium plays a key role in the immunogenicity of the trachea, epithelium denudation could reduce the immune response to tracheal allografts. In this way, we previously demonstrated the immune tolerance of epithelium-denuded cryopreserved tracheal allografts in a rabbit model [12], and we were able to achieve successful immunosuppressant-free OTT of short tracheal segments in this animal. Unfortunately, OTT with long tracheal segments failed because of luminal stenosis [13].

Therefore, a reliable method for the repair of large, fully circumferential tracheal defects remains to be found. We believe that the validation of our preclinical model of an epithelium-denuded cryopreserved tracheal allograft in a large animal is required before it can be considered for clinical applications. Because the pig is considered an excellent preclinical model for tracheal research, we first attempted a 1-stage OTT using either an autograft or an epithelium-denuded allograft wrapped with the strap muscles to achieve neoangiogenesis. Unfortunately, both animals died of graft necrosis on Day 6 and Day 7, respectively (unpublished data). Because the fate of the graft may vary according to graft length [9–11] and either heterotopic or orthotopic positions [12, 13], the goal of our present study was to evaluate the viability, immune tolerance and stiffness of a heterotopically transplanted tracheal allograft wrapped in the omentum in Landrace pigs. Preliminary results were reported briefly [14].

MATERIALS AND METHODS

The experimental protocol (APAFIS # 2963) received the approval of the French Ministry of Education and Research (Ministère de l’Enseignement et de la Recherche), provided that we limited the number of recipients to 2 groups of 5 animals each. Investigations were conducted in compliance with the experimental animal use guidelines of the French Ministry of Agriculture, Food-Processing Industry and Forest (Ministère de l’Agriculture, de l’Agroalimentaire et de la Forêt) that regulates animal experiments in France.

Study design

Donors and recipients were young Landrace pigs (Pannier SA, Wylder, France) weighing 26.8 ± 5.3 kg. Tracheae were retrieved from 6 male donors. Five tracheae were divided in 5 short (5-ring) and 5 long (10-ring) tracheal segments, which were denuded of their epithelium and carried to the tissue bank for cryopreservation. The remaining trachea was used as a control.

After thawing, the tracheal segments were heterotopically wrapped in the omentum in 2 groups of 5 immunosuppressant-free female recipients each.

Euthanization was scheduled from Day 6 to Day 90 in both groups.

Tracheal harvesting and cryopreservation

After premedication with intramuscular ketamine (10 mg/kg) and Sedaxylan® (2 mg/kg), donors were euthanized by an intracardiac bolus of embutramide, mebezonium and tetracaine (0.3 ml/kg) (T61; Intervet, Beaucouzé, France). The trachea from the cricoid arch to the right upper lobe bronchus originating from the trachea (a normal formation in the pig) was harvested via a midline cervical and trans-mediastinal route. Harvested tracheae displaying a truncated cone shape and sharing 20 of 22 tracheal rings were divided to provide 5 short (5-ring) and 5 long (10-ring) segments retrieved from the cranial and caudal parts of each trachea (Fig. 1). Consequently, the mean length of the short and long segments was 32.6 mm (SD = 2.4) and 55.4 (±3.5) mm, respectively. All segments were denuded of the epithelial layer by sharp dissection (Supplementary Material, Fig. S1A). Microscopic examination of a tracheal segment control demonstrated the effectiveness of the epithelium denudation (Supplementary Material, Fig. S2). Tracheal segments were carried to the tissue bank within 6 h of harvesting (European Homograft Bank, Brussels, Belgium) in cold Hank’s medium 199 (4–8°C) containing antimicrobial agents. After incubation in this solution for 48 h, cryopreservation was performed as previously described [12]. Briefly, the grafts were immersed in a cryoprotecting medium (10% dimethyl sulphoxide in Hank’s solution 199) before cryopreservation, then stored in the vapour phase of liquid nitrogen (−150 to −187°C). The grafts were stored in the tissue bank for 33–232 days (Table 1).

Tracheal heterotopic transplantation and follow-up

Before use, each tracheal segment was thawed as detailed previously [12]. Briefly, it was preserved at room temperature for 5–6 min, then placed in a water bath at 37–40°C for 8–10 min to reach a temperature of 4°C. The dimethyl sulphoxide was washed out with cold isotonic saline (+4°C) in 4 steps, decreasing its concentration from 10% to 6.6%, to 3.3% and to 1%. Finally, the tracheal segment was preserved in isotonic saline.

Recipients were 2 groups of 5 female Landrace pigs each. Premedication was similar to that performed in the donors. After orotracheal intubation, anaesthesia was provided via inhalation of isoflurane (2%) and oxygen (60%). With the animal in a supine position, a midline laparotomy was performed. Before use, the tracheal segment was stented with a 16-mm outer diameter silicone tube fixed by stitches (3/0 absorbable PDS II, Ethicon France, Issy Les Moulineaux, France) at both edges of the graft. The fixed silicone tube was long enough to maintain sufficient stretching of the intercartilaginous ligaments (Supplementary Material, Fig. S1B), thus facilitating neoangiogenesis from the omentum and reperfusion of the graft through the intercartilaginous spaces [8]. The distal part of the omentum was rolled around the graft and sutured with 3/0 absorbable PDS II sutures (Fig. 2A and B). The laparotomy was closed. Postoperative analgesia was provided with an intramuscular injection of buprenorphine 20 µg/kg on Day 1. No immunosuppressive drugs were given. Animals were housed in individual boxes and allowed food and water ad libitum. They were monitored and, after they were humanely killed, grafted tracheae were explanted on Days 6, 14, 30, 58 and 90 from each group.

Macroscopic and microscopic examination

After the silicone tube was removed, the grafts were evaluated as to morphology and size. The 10-ring grafted tracheae were divided to provide a 6-ring specimen reserved for microscopic examination and a 4-ring specimen for biomechanical tests. The 5-ring grafted tracheae were reserved for microscopic examination.

For microscopic assessment, the specimens were fixed in formalin for 2 days, embedded in paraffin and cut into 3-µm slides. They were stained with haematoxylin–eosin and saffron. Findings were compared with the morphology of both fresh and epithelium-denuded tracheal segment controls.

Lymphocyte infiltrate was scored semi-quantitatively on a scale of 1 to 3 as described by Shi and co-workers: 1 indicated minimal infiltrate (area of infiltrate <30% of the graft surface); 2 indicated moderate infiltrate (30% < area of infiltrate <70%); and 3 indicated severe infiltrate (area of infiltrate >70%) [15]. Chondrocyte viability was graded semi-quantitatively from 0 to 3, according to Nakanishi’s score, with 0 indicating severe changes (necrotic area of the cartilage surface >70%), 1 indicating moderate changes (30% < necrotic area <70%), 2 indicating mild changes (necrotic area <30%) and 3 indicating the absence of any abnormalities [16]. Cartilage calcifications were assessed as follows: +, mild, central deposits;  ++, moderate, 10%–50% of the cartilage surface;  +++, intense, superior to 50% of the cartilage surface, as detailed elsewhere [17]. Graft fibrosis was graded as follows: +, mild fibrosis of the lamina propria;  ++, moderate fibrosis thickening the lamina propria (less than 3 times as thick as the normal lamina propria) extending to the pericartilaginous tissue;  +++, severe fibrosis thickening the lamina propria (more than 3 times as thick as the normal lamina propria) extending to the pericartilaginous tissue.

Detection of apoptotic events

Aside from lymphocyte infiltrate, apoptosis is considered another, albeit non-specific, marker of rejection [18]. An immunohistochemical technique for detection of apoptotic cells permitted an additional assessment of immunological graft rejection, as demonstrated elsewhere [12, 13]. Briefly, apoptotic cells in the different components of the grafts were assessed with a murine monoclonal antibody (APOSTAIN), which is specific for fragmented single-stranded DNA, a hallmark of apoptosis.

Radial compressive force evaluation

Radial compressive force evaluation was performed on each 4-ring segment retrieved from the five 10-ring grafted tracheae. Tests similar to those previously described in a rabbit model [19] were performed on a tension–compression machine (Fig. 3A), equipped with a 2-kN (±0.01 N) load cell (INSTROM 5802). The 4-ring segments were placed between the compression grips of the machine, which were covered with surgical drapes (Fig 3B). They were subjected to uniaxial compression testing in the axis of the lateral diameter at a rate of 1 mm/min until 50% of the lumen occlusion was reached. The representative force (N) versus the displacement (millimetre) profile for a fresh 4-ring segment control obtained from yield load testing is shown in Supplementary Material, Fig. S3, which increases in a linear fashion until 50% lumen occlusion is reached. Load (N) versus displacement (millimetre) data were obtained for each 4-ring segment and compared with those obtained from a fresh tracheal segment control. Finally, the compressive radial load (N) at 50% lumen occlusion was calculated for each 4-ring segment.

RESULTS

All 10 pigs survived the operations, and none experienced any complications. Humane killings were performed sequentially on Days 6, 14, 30, 58 and 90 in both groups of 5 recipients.

Clinical, pathological and biomechanical findings are summarized in Table 1.

Macroscopic assessment

The 5-ring and 10-ring-grafted tracheae surrounded by the recipient’s omentum showed evidence of revascularization from Day 6 onwards. They were found to have a normal morphology with well-recognizable cartilage rings and membrane, with no significant reduction of the luminal diameter and length (Fig. 4). They showed no signs of suppuration or necrosis.

Histopathological findings

Grafted tracheae did not display any signs of acute rejection, which is characterized by capillary thrombosis and consecutive necrosis [8]. Lamina propria and pericartilaginous tissue exhibited significant collagenized fibrosis that increased over time. Increased neoangiogenesis, made up of small capillaries, involved the entire grafts, particularly the lamina propria (Fig. 5A), and was present in all types of tracheal allografts, regardless of length. A non-specific inflammatory reaction of neutrophils, eosinophils and macrophages was present and decreased over time. Similarly, lymphocyte infiltrate graded between 1 and 3 according to the Shi et al.’s score [15], was present up to Day 30, but in grafts studied after 58 days, this lymphocyte infiltrate was attenuated. Except for the graft transplanted in Pig 8, which exhibited vanishing chondrocytes, varying amounts of viable chondrocytes were observed in the cartilage rings of the other specimens. Islets of calcification were also present, increasing over time (Fig. 5B).

Detection of apoptotic cells

Of the 9 grafts studied, only the graft from Pig 8, which was transplanted with a 10-ring graft, displayed a small number of apoptotic cells within the cartilage tissue (Supplementary Material, Fig. S4). No apoptotic events were seen in the lamina propria and pericartilaginous tissue.

Compressive force resistance

A decrease in the radial force needed to occlude the lumen of 4-ring tracheal segments up to 50% was found early after the procedure (up to Day 30), but strength properties increased over time (Fig. 6). Biomechanical tests demonstrated that the representative stiffness required to achieve 50% lumen occlusion on Day 90 was similar to that of a fresh 4-ring tracheal segment (4.6 × 10⁻² vs 4.4 × 10⁻² N/mm) (Fig. 7).

DISCUSSION

In our porcine model, epithelium-denuded cryopreserved tracheal allografts heterotopically transplanted in the omentum showed satisfactory morphology and viability, regardless of graft length and acceptability, despite the absence of immunosuppressive drugs. Furthermore, they displayed sufficient biomechanical characteristics from Day 58.

A reliable method of OTT for the repair of large, fully circumferential tracheal defects is paramount in both benign and malignant tracheal diseases, but the absence of a well-defined blood supply to the graft and the requirement for immunosuppression constitute major limitations. Our previous work in a rabbit model demonstrated the feasibility of immunosuppressant-free OTT with short epithelium-denuded cryopreserved tracheal allografts but a failure in long transplants [13]. To resolve this issue, investigation of such allografts in a large animal model was required.

There is empirical evidence that the 2-stage technique with revascularization in heterotopy followed by OTT could be the procedure of choice [7, 8, 11]. Notably, Maksoud-Filho and co-workers assessed histological findings of heterotopic and orthotopic transplants wrapped in an omentum flap in immunosuppressed rabbits and demonstrated better graft neoangiogenesis in heterotopy rather than orthotopy [7]. Similar findings were shown by Li and co-workers using 6-ring tracheal autographs implanted into the omentum in a canine model [11]. Finally, Delaere and co-workers investigated indirect revascularization of heterotopically implanted tracheal allografts by means of a fascial flap wrap in immunosuppressed rabbits, allowing successful 2-stage OTT [8]. In the clinical setting, 1 patient had a 1-stage OTT with omentopexy. He sustained graft stenosis and consecutive permanent stenting [4]. On the other hand, another patient had a 2-stage OTT with revascularization of a tracheal allograft into the sternocleidomastoid muscle followed by OTT 3 weeks later. This patient had a satisfactory outcome [3].

Rejection is another major issue of OTT. Whereas the tracheal cartilage is well tolerated, the major histocompatibility complex Class II antigen, which is highly expressed in the epithelial layer, plays a key role in the immune response to tracheal allografts. Some studies have demonstrated a reduction of their immunogenicity using the cryopreservation process, which allows successful OTT of omental flap-wrapped tracheal allografts without immunosuppressive drugs in canine models [20, 21]. However, in our previous study in a rabbit model, it seemed to us that the cryopreservation process induced insufficient immunological tolerance, and we demonstrated the value of concomitant epithelium denudation/cryopreservation in reducing the immune response to tracheal allografts [12, 13]. Another benefit of epithelium denudation is the absence of mucus production inside the lumen of the heterotopically implanted graft [12]. Indeed, accumulation of mucus in a non-functional trachea enhances the development of purulence, which contributes to progressive graft necrosis [8].

Based on our previous experience in the field [22], we conducted our experiments in Landrace pigs. Because the lack of available fascia superficialis in this animal impedes graft wrap such as that performed in rabbit models [8, 12, 17, 23], we chose the omentum as a vascular carrier for graft revascularization. Our investigation demonstrated effective neoangiogenesis of grafted tracheae regardless of length. Given that grafted tracheae retrieved on Day 6 showed high levels of neocapillaries, we expect that neoangiogenesis occurred early after implantation. Additionally, grafts displayed fibrosis of the lamina propria and pericartilaginous tissue, most likely resulting from the critical ischaemic period preceding effective neoangiogenesis. In the present study, epithelium-denuded cryopreserved tracheal allografts heterotopically transplanted displayed moderate signs of rejection, which were attenuated over time, with preservation of the graft architecture and cartilage viability. This immunological tolerance was further confirmed by the absence of apoptotic events in the majority of the studied allografts.

In the field of experimental tracheal surgery, some biomechanical studies analysing the tensile strength of the porcine trachea until rupture have been reported. They were conducted to assess the properties of a tissue-engineered tracheal matrix [24] or the tensile strength of tracheal anastomoses [25]. However, they did not address the question of the ability of the tracheal graft to resist respiratory collapse. Recently, Hoffman and co-workers reported a characterization of the biomechanical properties of the porcine trachea, but biomechanical tests for radial compressive force evaluation were lacking [26]. To our knowledge, the present study is the first investigating this parameter in the pig model, demonstrating an increase in strength from Day 58, likely because of cartilage calcification over time. Therefore, a minimum 2-month threshold would be required before considering transposition of the graft in orthotopy.

Further investigations are required to assess the reliability of the epithelium-denuded cryopreserved tracheal allograft for OTT, mainly in terms of strain abilities. Given that the respiratory epithelium provides a barrier against microbiological contamination, there is a real risk for suppuration of the grafted trachea, as evidenced in our previous study in a rabbit model [13]. In that study, we also had to deal with the possibility of luminal narrowing due to fibroblastic development arising from the lamina propria, which was, however, prevented by concomitant re-epithelialization arising from the edges of the native trachea [13]. This issue could be resolved using a temporary silicone stent, thereby limiting stenosis until re-epithelialization of the airway surface is achieved, thus ensuring full functionality of the grafted trachea in terms of mucus production and clearance.

Finally, the omentum has been regarded as relevant for heterotopic revascularization of a human tracheal allograft [27]. Given that the omentum has the potential to reach the tracheal region, a 2-stage OTT with a graft wrapped in an omentum flap could be a feasible procedure in the clinical setting.

Limitations

In the present study, the lack of a control group might be considered an issue. However, the failure of tracheal allografts without epithelium denudation implanted in heterotopy in the absence of immunosuppressive drugs has been previously demonstrated, notably in the rabbit model [8]. According to the 3Rs rule (reduce, refine and replace), our local ethic committee for animal research did not allow us to use animals for confirmation of previously established data.

In clinical practice, the omentum is not always available. Thus, it is necessary to define other candidates as vascular carriers for graft revascularization. A cadaveric trachea wrapped with either trapezius or latissimus dorsi muscle flap has already been assessed for its ability to reach the tracheal region [28]. The pectoralis major muscle flap, which has similar rotational capability [1], might be another option. Another approach might be heterotopic revascularization of the tracheal allograft by means of a flap wrap into the fascia lata or a de-epithelialized anterolateral thigh flap and subsequent orthotopic transposition of the wrapped graft as a free flap procedure.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

ACKNOWLEDGEMENTS

We are grateful to Arnold Dive, Martin Fourdrinier and Michel Pottier for their contributions to this work.

Funding

This work was supported by the Fondation de l’Avenir, Paris, France [AP-RMA-2015-001].

Conflict of interest: none declared.

REFERENCES