Regenerative Response 35 Days After Esophageal Replacement in a Porcine Model; Technical Difficulties and Attempts to Achieve Optimal Tissue Remodeling

ABSTRACT

Background

In previous articles, a porcine model for bridging circumferential defects in the intrathoracic esophagus was developed. The aims of this present study were to evaluate the continued healing response after 35 days, avoid stent migration of the esophageal stent, and to investigate whether it would be beneficial to add new extracellular matrix (ECM) to the healing area after 20 days.

Methods

Surgery was performed in twelve piglets, and five different types of stents were used. In two piglets, new ECM was added by endoscope to the area of healing after 20 days. After the animals were euthanized, the esophageal tissue was examined.

Results

Histologic examination after 35 days showed clusters of desmin‐positive smooth muscle cells and the sprouting of nerves in the area that was healing. Generally, there were fewer M1 classically activated macrophages in specimens after 35 days when we compared them with the 20‐day study. The CD 163 positive macrophages (M2‐macrophages) were seen in all specimens.

Four piglets did not survive to the end of the study period because of adverse events. Out of the eight piglets that were euthanized after 34 to 35 days, six had stents that had migrated to the stomach. Only in two piglets, who had large rilled stents, did the stents remain in place throughout the study period.

Conclusion

After 35 days, the area of healing did not show more signs of regenerative healing than the 20 days study.

A procedure to add a new biomatrix by re‐stenting endoscopically after 20 days was performed on two pigs. The procedure was feasible, but due to limb pain in the animals, they had to be euthanized prior to the plan, which prevented the evaluation of the effect on the regenerative response.

The regenerative healing that was started needs to be further orchestrated in other ways to produce a more functional outcome with time.

1. Introduction

Different types of scaffolds have been used in experimental animal models to try to manage esophageal defects [1, 2]. Esophageal Atresia (EA) is a malformation with a congenital defect in the esophagus that affects approximately 1 in 2500 to 3000 births. Treatment of esophageal atresia is difficult and gives poor functional outcomes if more esophageal tissue is missing, so‐called Long Gap Esophageal Atresia (LGEA) [3, 4, 5, 6]. To compensate for the lack of esophageal tissue, we developed a porcine guided tissue regeneration model for studying esophageal regrowth after the removal of 3 cm of the intrathoracic esophagus [7]. The use of the body as a place for in vivo tissue engineering, relying on the body as a natural bioreactor, may have advantages for the development of more complex tissue. Extracellular matrix (ECM) scaffolds have proven to be effective for the reconstruction of small patch defects in the animal and human esophagus [8, 9]. Scaffolds have also been used for full circumferential defects in the human esophagus [10] but there is no approved product for this kind of use.

Previous animal studies of circumferential resection of the esophagus have always resulted in lethal strictures except when a stent has been used for at least 4 weeks [11, 12, 13]. In dogs, the esophagus is composed of skeletal muscle, and in humans, the esophagus is composed of both skeletal muscle and smooth muscle. It has been suggested that animals with skeletal muscle in the esophagus have better healing than the mixed‐type esophagus with smooth muscle. Skeletal muscle cells have satellite cells (SC:s) that can divide and form muscle fibrils [14]. Smooth muscle cells have no precursor cells and have to divide to form conglomerates that have the capacity to arrange motor function in the esophagus.

Pigs have a mixed type of esophagus. Therefore, the model with piglets was chosen to try to mimic the esophagus of the growing child.

To improve healing, different treatments have been developed to help the body form functional tissue after injury [1]. One of these treatments is the use of ECM, a collagen scaffold derived from different tissues in different species. There is extensive literature describing differences among ECM materials after placement in tissues in both experimental models and in humans [2]. Recent studies have shown that the ECM induces a healing response where the macrophages in the wound have another phenotype, the M2‐macrophage. Their presence in the wound allows more functional tissue to form (angiogenesis, amounts of organized smooth muscle cells and and ingrowth of nerves) [15, 16, 17, 18]. Macrophage activation is a complex field of research, where a continuum of different phenotypes has been identified, depending on the means of activation. The CD163 antigen is one of the surface markers that is specific for the remodeling M2 macrophages. M2 macrophages in wound healing seem to be one requisite of functional tissue in regenerative models, whereas M1‐macrophages seem to be the default macrophage after damage ending up in scar formation. The importance of macrophages for a successful regenerative response in lower vertebrates has been elegantly demonstrated in experiments performed on salamanders [19].

In our previous studies, an ECM bridging graft composed of a Biodesign (Cook Medical, Bloomington, IN, USA) sewn around a silicone stent was used. Biodesign is well known for clinical use in humans and can be purchased off the shelf. The area of remodeling was harvested after 9 and 20 days, and results were presented in two articles [20, 21], and we now wanted to study the healing response after 35 days. In the previous articles, we showed that leakage from the esophageal anastomosis gave rise to a more aggressive inflammatory pattern 8–10 days after replacement in our model [20]. We also saw that if the anastomoses were patent, the bridging area showed signs of regenerative healing with smooth muscle cells. In the recent study, the piglets were euthanized after 20 postoperative days, and the histologic result studied. If we had stent loss prior to 20 days, the results were affected and showed fewer CD 163 positive M2‐macrophages and fewer desmin‐positive smooth muscle cells. Stent migration is a known complication when stents are placed in the esophagus. The stent needs to have anti‐migrative properties to avoid the risk that it dislodges and moves down into the stomach. Stents can be flared at the ends, large in diameter, and have fins or scales. In this study, we wanted to examine the continued healing process 35 days after surgery. Since stent loss seemed to affect the healing response negatively, we also wanted to try to overcome the problem of stent migration.

The aims were first to describe the clinical outcome and histologic result 35 days after esophageal replacement, to investigate whether stent migration could be avoided, and to study how premature stent loss affects clinical and histological outcomes. Moreover, to determine whether reapplication of ECM 20 days after surgery could enhance an M2‐macrophage‐guided regenerative healing pattern 35 days after surgery.

2. Study Design: (Material and Method)

The study was approved by the Regional Ethical Committee for animal research in the west of Sweden, Dnr 197–2011, 2011‐06‐14. + 5.8.18–17 605/2017.

The study was performed on female newly weaned Sus Scrofa , 10 weeks of age and around 15 kg. Females were chosen since they are easier to care for after surgery. We operated on two animals in each instance, with the same method and with Biodesign. We had to operate on two animals so that they could be cared for together in the postoperative period. This is according to our ethical permit. After each pair, the problems achieving 35 days of successful stenting were addressed. Due to the migration of the stents, we changed the stent properties after each piglet pair.

The number of animals in each group was too small to draw conclusions about histological differences in pairs. It can, however, show which surgical‐technical methods could work for longer follow‐up times for regenerative studies in the future.

2.1. Surgical Procedure

Twelve weaned piglets (domestic pigs, Sus Scrofa Domestica) were operated on according to our previously described method [5]. Briefly, a 3‐cm long resection of the intra‐thoracic esophagus was performed through a right‐sided thoracotomy. In previous studies we used a 10 cm straight silicone stent. Around it a 4 × 7 cm 4‐ply Biodesign (Cook Medical, Bloomington, IN, USA) mesh was sutured as a graft to bridge the gap. The mucosa was closed with a running suture 5 mm from the edges of the mesh using 5–0 Biosyn. The musculature was closed 2 cm from the edges in order to provide a large zone of contact between the esophageal musculature and the Biodesign mesh.

In the first two pigs, we used the 10 cm straight silicone stent (ESSKA, Esska of Sweden AB, Linköping, Sweden) as was used in the previous 20‐day study [21] (Table 1). Due to stent migration after 24 and 30 days, we decided to change the stent's properties and dimensions.

TABLE 1.

Animal, Type of stent, Time of stent‐migration and sacrifice. [Color table can be viewed at wileyonlinelibrary.com]

Piglet	Stents		Post‐op stent migration (days)	Days at sacrifice
1 + 2	10 mm diam, 100 mm Silicone tube, ESSKA		30 + 24	35 + 35
3 + 4	12‐10‐12 mm diam, 100 mm Metal Bile stent, Boston Scientific		33 + 26	35 + 30
5 + 6	12‐10‐12 mm diam, 100 mm, degradable, Ella‐cs		30 + 33	35 + 35
7 + 8	25‐15‐25 mm diam, 110 mm Silicone stent, flared, rilled, Lue Engineering		No migration	35 + 34
9 + 10	25‐15‐25 diam, 110 mm Silicone stent, flared, Lue Engineering		P 9: 33 P 10: Died during primary surgery	34 + 0
11 + 12	12‐10‐12 mm diam, 100 mm Metal Bile stent, Boston Scientific		Stent in place at replacement after 20 days 20	20 + 29
	+ 17‐12‐15 mm diam, 100 mm Alimaxx metal stent, Merit Medical		Stent in place at sacrifice after 29 days

Piglets 3 and 4 were stented with a metal bile stent (Boston Scientific, Marlborough, Massachusetts, USA). This stent occluded and migrated into the stomach. It seemed like the lumen was too small, so the gruel got stuck and the stent thereafter migrated.

In piglets 5 and 6, a degradable stent was used according to our specifications (Ella‐cs, Hradec Kralove, Czech Republic). The degradable stent also became occluded and migrated into the stomach.

Piglets 7 and 8 had a stent with specific anti‐migrative properties, a large and flared silicone stent with ridges (Lue Engineering, Kungsbacka, Sweden). This stent was kept in place for the 35‐day period, but the ridges caused a star‐shaped pattern in the remodeled area, which is not an anatomical appearance (Figure 2a,b).

FIGURE 2.

(a) Showing imprints of the ridges from the silicone stent. (b) Showing imprints of the ridges of the silicone stent, higher resolution. [Color figure can be viewed at wileyonlinelibrary.com]

Piglets 9 and 10 were stented with a large and flared silicone stent without ridges (Lue Engineering). One of the pigs died of cardiac arrest during implantation in the esophagus, probably because of an extreme vagal reaction when the large stent was pushed in place. Though in the histologic investigation after 35 days, we saw that it did not look like more functional tissue. Therefore, in the next two piglets, we made a new plan to add more ECM to the healing area after 20 days to stimulate further regenerative growth.

Piglets 11 and 12 received a covered bile stent, 10 cm (Boston Scientific ). After 20 days, the piglets were anesthestized and endoscopic removal of the stent was performed. A Allimaxx (Merit Medical Systems Inc., South Jordan, Utah, USA) stent with a Biodesign‐mesh (Figure 1) was placed during endoscopy in the area of remodeling.

FIGURE 1.

Stent covered with Biodesign for endoscopic delivery after 20 days. [Color figure can be viewed at wileyonlinelibrary.com]

All piglets had a gastrostomy placed through a small midline laparotomy. A 24 Ch Nelaton (B.Braun Stockholm AB, Danderyd, Sweden) catheter was tunneled into the left side of the animal and placed in the stomach. The balloon was filled with 20 mL sterile water. The catheter was fixed with two purse‐string sutures that were tied and then fixed to the abdominal wall where the catheter left the abdominal cavity. Another purse‐string suture was applied to the skin. The pigs were given i.v. Tazocin (Pfizer AB, Stockholm, Sweden) 1 g b.i.d. (two times a day) on the operating day and then received enteral Bactrim (Eumedia Pharmaceuticals Gmbh, Lörrach, Germany) 8 mL b.i.d. until euthanized.

2.2. Postoperative Routines

Contrast x‐ray of the esophagus was performed if symptoms of dysphagia arose.

After surgery, the pigs were accommodated two to a room for ethical reasons. There was a grating between them to prevent them from damaging each other's gastrostomy. There was no bedding in their room to avoid problems with bedding material obstructing the bridging graft.

The piglets were fed with probe formula through the gastrostomy four or five times a day and were given water orally ad libitum. The formula was composed of Infantrini junior (Danone AB, Solna, Sweden) Resource energy (Nestlé Sverige AB, Stockholm, Sweden) Resource protein (Nestlé Sverige AB, Stockholm, Sweden) and rapeseed oil at each meal. The composition was formulated to approximate the nutrients in commercial pig food. The gastrostomy was removed on post‐operative day 10, and the formula was given orally.

To prevent them from suffering from reflux, which also might affect the healing of the bridging area, all piglets were given a proton‐pump inhibitor in the form of Nexium (Grunenthal Sweden AB, Solna, Sweden) or Lanzo (Pfizer AB, Stockholm, Sweden) orally.

Contrast x‐ray of the esophagus was performed by feeding contrast‐enhanced food to the animals in a wooden box that could be fitted into a fluoroscope.

2.3. Harvesting Procedure

The piglets were planned for sacrifice 34 or 35 days after surgery. They were sedated using Zolitil (VIRRBAC, Carros, France) and Domitor Vet (Orion Pharma AB, Danderyd, Sverige) and euthanized using Alfatal vet (Omnidea AB, Stockholm, Sweden).

The bridged area was harvested for histological analyses. The specimen was divided into proximal and distal halves, and the proximal part was marked with a suture. In all pigs, a circumferential sample was also taken from the center of the bridged area.

2.4. Histological Method: (Table 2)

TABLE 2.

Cells/extracellular components stained and antibody markers.

Cells/extracellular component stained	Histological marker
Pericyte, smooth muscle cell, skeletal muscle cell	Desmin
Smooth muscle cells, fibroblasts	Smooth muscle Actin
Smooth muscle cells	Smooth muscle Myocin
Basal Membrane	Laminin
T‐lymphocytes	CD3
M1‐macrophages, neutrophil granulocytes	Calprotectin
M2‐macrophages	CD 163
Small synaptic vesicles, nerve specific	Synaptophysin

Samples were fixed in 4% buffered formaldehyde, dehydrated, cleared, and embedded in paraffin. Four‐micron thick longitudinal or transverse sections were cut. The sections were deparaffinized, hydrated, and treated with a solution of 3% hydrogen peroxide to quench endogenous peroxidase activity. After heat‐induced antigen retrieval in an EDTA buffer, pH 8.0, the sections were rinsed and blocked with 2.5% horse serum (Vector Laboratories, Burlingame, CA) for 30 min. The sections were incubated overnight at 5°C with monoclonal mouse antibodies raised against desmin (DAKO, Glostrup, Denmark), smooth muscle actin (SMA) (DAKO) and smooth muscle myosin (SMM) (Sigma‐Aldrich, Saint Louis, Missouri, USA). To estimate the extent of inflammatory reaction, further sections were incubated with a mouse monoclonal antibody staining calprotectin that recognized neutrophils, monocytes, and classically activated (M1) macrophages (DAKO). A polyclonal antibody against CD163 (Antibodies in line, Aachen, Germany) recognized a subset of macrophages of M2 phenotype. Further sections were incubated with monoclonal mouse antibodies (DAKO) to neurofilament (NF), recognizing the cytoskeleton of nerve cells. Anti‐rabbit Impress Reagent HRP or anti‐mouse Impress Reagent HRP (Vector Laboratories, Burlingame, CA, USA) were used as secondary reagents, and the immunoreactions were visualized using liquid DAB+ substrate (DAKO). Nuclei were counterstained with hematoxylin. The sections were then dehydrated and mounted using DPX (Merck Millipore, Darmstadt, Germany).

3. Results

3.1. Clinical Results

Eight out of twelve animals were sacrificed on days 34 or 35 according to plan. One piglet died during surgery and three piglets were euthanized prior to plan (Table 1).

An esophageal X‐ray was done two times in animals 3,4,5,6 and 9 after about days 20 to 23 and days 29 to 35. We started to do contrast x‐ray in piglets 3 and 4 to find signs of stent loss in case of dysphagia. Piglets 7 and 8 did not have signs of dysphagia and therefore, an x‐ray was not performed. In piglets 11 and 12, we did the endoscopic procedure after 20 days; therefore, they were not suitable for fluoroscopic examination.

In all animals, the harvested bridged area was seen as a full tissue tube, white‐red in color and with an approximate length of 1.7 to 2.0 cm.

In piglets 1 and 2, we used the 10 cm straight silicone stent as was used in the nine‐and twenty‐day studies [20, 21]. They had symptoms of dysphagia after about 30 days, and in both, the stent had migrated at sacrifice. To try to avoid stent migration, we changed the stent properties.

A metal bile stent was used in animals 3 and 4. The stents were in place at the first esophageal X‐ray, but both subsequently migrated and were occluded by food at sacrifice. Pig 3 started to eat more slowly two days before sacrifice. Piglet 4 had symptoms of severe dysphagia and was sacrificed prior to plan after 30 days. On day 23, the X‐ray showed that the stent was in place, but it had moved on day 29 when the symptoms of stricture were severe.

In piglets 5 and 6, a degradable stent was used. Piglet 5 showed signs of dysphagia after 28 days, and an X‐ray five days before sacrifice showed that the stent had migrated. Piglet 6 ate without problems, and the stent was in place on the X‐ray. At sacrifice, both stents had migrated into the stomach.

Piglets 7 and 8 had the stent in place at 35 days. The stent was the large and flared silicone stent with ridges, which seemed to have the best anti‐migrative properties. The piglets showed no clinical signs of stricture formation and could eat without problems. In the bridged remodeled area, there were deep imprints of the ridges of the stent. The stent seemed to have good anti‐migrative properties, but the ridges affected the normal appearance of the esophageal tissue layers (Figure 2a,b).

A large and flared silicone stent was used in piglets 9 and 10. Piglet 9 ate well until sacrifice, and the X‐ray revealed no migration, but the stent was found in the stomach at sacrifice. Unfortunately, pig 10 died day 1 during the endoscopic placement of the large stent at the primary surgery.

Piglets 11 and 12 were planned for re‐stenting with Biodesign after 20 days since previous work showed that the stents did not migrate before 20 days. The hypothesis is that adding new biomatrix would enhance the remodeling inflammatory response to achieve more functional tissue. Piglets 11 and 12 received a metal bile stent with Biodesign at the first surgery, and after 20 days, the stent was removed. An Alimaxx stent covered with a Biodesign sheet was endoscopically introduced and placed in the bridging area (Figure 1). Piglets 11 and 12 had no problem eating, but due to unexpectedly severe limb pain in the animals, they had to be sacrificed prior to plan at 20 and 29 days.

3.2. Histological Results

Samples from all pigs, except piglet 10 that died during primary surgery, were processed according to the histological method previously described. In pigs 1 to 9 the mesh was almost entirely degraded and only traces of Biodesign could be seen, while in pig 12 most of the mesh put in place 20 days after the operation was still present. In the bridging area, desmin‐positive smooth muscle cells appeared in clusters (Figure 3), abundant in some specimens, sparse in others. There was a typical transition zone of about two to three mm from the muscle edges, where dense bundles of smooth muscle cells spread into the bridged area (Figure 4). In some samples, inflammatory, calprotectin‐positive cells could be seen in islets covering most of the bridging area.

FIGURE 3.

Higher magnification, desmin staining showing small clusters of muscle cells. [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 4.

Transition zone from native esophagus to healing area. Desmin staining. [Color figure can be viewed at wileyonlinelibrary.com]

In pigs 1, 5 to 9, and 11 and 12, there were nerve fibers sprouting into the bridged area when staining for NF (Figure 5a,b). All fibers seemed to be adjacent and sprouting near the vagal nerve.

FIGURE 5.

(a) Nerve fibers sprouting into the bridged area. NF staining. (b) Nerve fibers sprouting into the bridged area, higher resolution. [Color figure can be viewed at wileyonlinelibrary.com]

With staining for CD163 positive cells, M2‐macrophages were seen in all samples but were sparse and located close to the esophageal lumen. There was no clear correlation between CD 163 positive cells and histological tissue reformation in the samples. The mucosa could be seen covering the gap in some samples. However, it seemed loosely attached after removal and histological processing, as if it had been peeled off (Figure 6).

FIGURE 6.

Mucosa loosely attached, after removal and histological processing. [Color figure can be viewed at wileyonlinelibrary.com]

The histologic result is summarized in Table 3. Stereological analysis through random distribution of the analyzed parts of the specimens was not possible to do. Other articles with a large animal model have the same problems. Therefore, a semi‐quantitative scale was used.

TABLE 3.

The grading − to ++++ is a semi‐quantitative scale, where the histological slides from the resected area with newly formed tissue have been analyzed and estimated. In the muscle growth/reformation of tissue column, the best areas of remodeling have been compared between the slides.

Resected area of esophagus
Muscle growth/reformation				Inflammatory pattern (Calprotection/CD163)		Stent migration (days post‐op)
Piglet no	Desmin	Myosin	Neurofilament (NF)	Calprotectin	CD 163
1	+	+	+	+++	+	30
2	++	+	−	++	+	24
3	−(+)	−	−	++++	+++	33
4	−	−	−	+++	++++	26
5	+++	+++	+	+	+	30
6	++++	+++	+	+	+	33
7	+++	+++	++++	+	++	No migration
8	+	+	++++	+++	+++	No migration
9	+	+	+++	+++	+	33 days
10	Died during primary surgery
11	++++	++	+++	++++	++++	No migration
12	+	++	++	++	+++	No migration

The histological slides from the resected area with newly formed tissue have been analyzed and estimated. In the muscle growth/reformation of tissue column, the best areas of remodeling have been compared between the slides.

Generally, in this study we found fewer M1 classically activated macrophages in specimens than in the 20‐day study, independent of stent loss prior to sacrifice.

4. Discussion

This study showed that there were multiple problems to achieve continued remodeling of the esophagus between 20 and 35 days after esophageal replacement with a bridging graft composed of a stented Biodesign. Methods to avoid stent migration in the model were developed. Earlier work with esophageal replacement has shown that stricture occurred unless the area had been stented for at least 30 days [11, 12, 13]. Our previous studies showed promising results up to 20 days with increased amounts of smooth muscle cells, angiogenesis, and ingrowth of nerves.

Using a large and flared stent, the anti‐migrative properties kept the stent in place for 35 days. These large stents, however, made the surgery and implantation more difficult. One piglet died during the replacement of the large stent, probably because of extreme vagal stimulation when the stent was endoscopically pushed through the mouth and upper esophagus. The problem with stent migration is model‐related, as the piglets grow to at least double their size during the time period of 35 days. In an infant or an adult human, this would be less of a problem since humans grow slowly until they become adults. In the piglet model, we suspected that it would be better to endoscopically change the stent after a time period of 20 days. This was to avoid more difficult surgery during the first implantation. The hypothesis was also that adding new ECM after 20 days with restenting Biodesign to the healing area could promote the development of more functional tissue and recruit a higher number of smooth muscle cells and nerve fibers, as well as fewer fibroblasts. The endoscopic procedure with restenting and adding more Biodesign was feasible. Unfortunately, the piglets had to be sacrificed prior to plan because of limb pain. We think that the lameness was coincidental. Lameness in pigs is a common problem in domestic pigs. Some studies have shown that growing pigs have problems with limb pain and have to be euthanized [22]. Selenium deficiency is known to cause limb pain, but adding extra selenium did not help these pigs. Another cause of lameness is the increasing rate of infection with Mycoplasma (M.) hyosynoviae [23]. Thus is lameness is common in Sus Scrofa when they grow, and we did not find any surgical‐related cause for the lameness in our study.

The histological result at sacrifice after 34 and 35 days was similar to the previous 20‐day study [21] with a sparse layer of smooth muscle cells in the area of healing. At 29 days (eight days after restenting and adding new Biodesign) the histology did not show a greater number of smooth muscle cells, but it did show high numbers of M2‐macrophages.

Generally, in this study we found fewer M1 classically activated macrophages in specimens than in the 20‐day study, independent of stent loss prior to sacrifice. This, together with stagnation of progress in the regenerative type of healing after 35 days, could imply that the ECM material shown to enhance M2‐macrophages might be lacking after 20 days.

The CD 163 positive macrophages (M2‐macrophages) were seen in all specimens (Table 3). The 20‐day study showed CD163 positive macrophages only when the stent was in place, but this finding could not be seen after 35 days. Our previous hypothesis, that M2‐macrophages inhibit the M1‐driven inflammation and result in less scarring [21], was not supported by these findings.

In the last piglets in the series, piglet 11 and 12, there were higher amounts of CD 163 positive macrophages, suggesting that this was enhanced by ECM presence.

In the field of tissue engineering, there have been many types of in vitro alternatives, where a bioreactor is used to simulate the in vivo conditions. Studies like this, relying on the use of the body as a natural bioreactor, may have advantages as the regenerated tissues could be more functionally integrated by being formed in situ. The aim is to enhance the inflammatory pattern and regeneration so the regenerated organs will be more functional. A study published by Francesca et al. 2018 [24], replacing the mid‐thoracic porcine esophagus with a synthetic scaffold carrying autogenous stem cells, showed promising numbers of smooth muscle cells in the formed esophageal tube. However, compared with native esophageal tissue, there was still a low number of smooth muscle cells. In the study by La Francesca et al. [24], Platelet rich plasma (PRP) and aMSC (aidipose Mesenchymal Stem Cells) were applied to enhance the regenerative type of inflammation.

The histological appearance in our model at 35 days is similar to the results of La Francesca, et al. As no evident progress in the abundance of muscle and nerve ingrowth is seen from day 20, there is no reason to believe that the replaced segment would continue to remodel without further stimulation. The results from our previous 20‐day study and this 35‐day study indicate that optimization in the early healing response is needed to generate more functional tissue. If more functional tissue with a similar anatomical appearance as the human esophagus is achieved, it could be used to bridge circumferential defects in humans in the future.

We believe that many biological aspects of the immune system must be orchestrated to promote continued regenerative healing at the right time. That might include adding stem cells, PRP, ECM, or specific tissue growth factors. Another aspect is the role of reinnervation in the reformation of functional tissue after resection [25, 26]. In our study, NF‐stained sprouting nerve fibers could be seen close to the vagal nerve (Figure 5a,b), but we do not know whether nerve fibers will continue to grow towards the resected area to promote signals and control of the reformed tissue.

5. Limitations

Limb pain/lameness is a common problem with domestic pigs [22, 23]. They grow very fast and gain weight at about 0.5 to 1 kg per day. The last two piglets were sacrificed earlier to plan because of lameness, so we could not investigate how the tissues would have developed by adding Biodesign after 20 days. Therefore, limb pain is a limiting factor in our model. Studies on pigs are often done with Yucatan mini pigs [24]. They do not grow as fast, do not have problems with lameness, and might be a better alternative in a future study.

Another limitation of this study is the small sample size. As we work with a cumbersome model, we try to use as few animals as possible in each step to find which factors are important to achieve regenerative healing.

6. Conclusion

A porcine model for bridging circumferential defects with stented ECM scaffold in the intrathoracic esophagus was developed, in which a healing response that took place over 35 days was evaluated. It was found that stent migration could be avoided by using a large stent with anti‐migrative properties, but this made the surgery difficult, and there was a risk of adverse events for the animals. Premature stent loss correlated to strictures and a poor histologic result. The reapplication of new ECM scaffold (Biodesign) endoscopically was feasible and promising, but unfortunately, the piglets had to be sacrificed prior to plan as they developed an adverse lameness that is quite common in domestic pigs.

The histology was compared to a previous study where the healing response was evaluated after 20 days. After 34 to 35 days, the granulation tissue contained fewer M1 classically activated macrophages than the 20‐day study, similar numbers of CD 163 positive macrophages (M2‐macrophages) and similar numbers of newly formed smooth muscle cells. Therefore, we believe that the regenerative response that was started needs to be orchestrated in other ways to produce more functional tissue. Further studies are needed to investigate the possibilities to improve the regenerative response.

Author Contributions

A. Sandin: concept/design, data analysis/interpretation, drafting article, data collection. L. Jönsson: concept/design, data analysis/interpretation, critical revision of article. M. Dellenmark Blom: critical revision of article. L.‐G. Friberg: concept/design, critical revision of article. V. Gatzinsky: concept/design, critical revision of article. E. Jennische: critical revision of article, data collection, data analysis/interpretation, funding. K. Abrahamsson: critical revision of article, data collection, funding, approval of article. O. Holmqvist: data analysis/iterpretation, data collection, critical revision of article.

Conflicts of Interest

The authors declare no conflicts of interest.

Sandin A., Jönsson L., Jennische E., et al., “Regenerative Response 35 Days After Esophageal Replacement in a Porcine Model; Technical Difficulties and Attempts to Achieve Optimal Tissue Remodeling,” Artificial Organs 49, no. 8 (2025): 1288–1297, 10.1111/aor.15006.