Esophageal cancer is a common malignant tumor of the digestive tract (Qiu et al., 2021; Sung et al., 2021). With advancements in endoscopic technologies, more esophageal diseases can be diagnosed and treated through endoscopy (Rabinowitz et al., 2020a; Rabinowitz et al., 2020b; Tang et al., 2020; Hoibian et al., 2021). Endoscopic submucosal dissection (ESD) has become the preferred therapy for early intramucosal esophageal cancers (Probst et al., 2015; Qi et al., 2018), especially because the R0 resection rates are higher. However, the cardinal long-term complication of ESD is esophageal stricture (Mizuta et al., 2009; Isomoto et al., 2013). In particular, an esophageal mucosal defect occupying more than three-quarters of the lesion circumference is an independent risk factor for esophageal stricture after ESD (Tang et al., 2021; Zhang et al., 2021). It is worth noting that almost all patients who undergo full circumferential ESD eventually develop esophageal stenosis (Yu et al., 2019). Currently, the rate of esophageal stricture prevention is unsatisfactory (Bhatt and Mehta 2020; Zhang et al., 2020).

With tissue engineering and regenerative medicine approaches, including the use of biologic scaffolds or cell-based therapies, researchers are exploring treatments that can reconstruct structurally and functionally normal tissues. One of such novel treatments comprises the use of an acellular dermal matrix (ADM)—a type of bioresorbable membrane—that showed effectiveness in previous animal experiments for providing local coverage of the mucosal defect in preventing esophageal stenosis after hemi-circumferential ESD (Han et al., 2017). Nevertheless, stenosis after full circumferential ESD requires viable solutions, and the effectiveness of ADM in circumferential ESD could not be verified until now. The application of an ADM in full circumferential ESD is challenging, because ADM fixation in a circumferential excision wound is difficult.

Therefore, we developed a new type of fully-covered self-expanding metal stents (FCSEMS) with an attached ADM and aimed to evaluate its feasibility of application and efficacy for esophageal stenosis prevention after full circumferential ESD in a porcine model.

Prior to conducting our formal experiment, we first demonstrated that in the absence of intervention, the mucosal defect formed by 1-cm circumferential esophageal ESD could lead to esophageal stenosis in a porcine model. All the mini pigs developed esophageal stricture without the ability to tolerate oral intake, thus requiring euthanasia between postoperative weeks 2 and 3. On endoscopy, we found pinhole strictures and an obvious absence of epithelial tissue at the ESD site (Supplementary Figures S2A–S2B). Moreover, the pathological assessment of the esophageal specimens revealed a lack of continuous epithelium accompanied by significant submucosal fibrosis (Supplementary Figures S2C–S2D), suggesting that the establishment of full-circumference ESD by the above method can lead to esophageal stenosis without any intervention.

The allocation of the 12 minipigs was random between both the groups. There were no significant differences in the dysphagia score and stenosis 2 weeks after ESD between the two groups. At week 4 post-ESD, the animals in group A had lesser dysphagia score than those in group B. The reduction in the luminal circumference of the esophagus in group A was lower than that in group B, which was in line with the endoscopic results. The macroscopic evaluation indicated that ADM could decrease esophageal mucosal contracture by approximately 40%.

In group A, the pathological analyses revealed a well-regulated neo-epithelium, lesser submucosal layer fibrosis and inflammatory infiltration, and greater re-epithelization rate than those in group B. These findings could be explained by the efficacy of the ADM patch. One possible explanation is that the matrix can trap and bind the recipient’s own cells and growth factors, which favors the growth of esophageal epithelial cells and vascularization. Further, this can help achieve rapid epithelialization in the mucosal defect (Rennert et al., 2013; Xu et al., 2021). Simultaneously, in the wound area, the matrix can inhibit the fibrotic formation and infiltration towards the deeper muscular layer, which can effectively prevent muscularis propria injury and is important for tissue repair and reconstruction.

Since the mean CRP rise in group A was lower than that in group B at week 4 after ESD, the covering of mucosal defects by ADM may have an anti-inflammatory effect in circulation. The degree of CRP level increase is positively correlated with the degree of infection. Hence, from our comprehensive analysis of the local inflammatory cell counts and general elevation in CRP levels, the ADM patch was deemed capable of attenuating the inflammatory infiltration in the local defect and inflammatory reaction in the circulation. ADM has several potential advantages relating to its source tissue and manufacturing process, which may have contributed to the limited inflammatory response observed in this study. ADM is manufactured with all lipids, fats, carbohydrates, and non-collagen removed, thus retaining only collagen. This can minimize the inflammatory reactions and host immune response, resulting in less tissue damage due to inflammatory cell aggregation (Cornwell et al., 2009; Lantis et al., 2021).

Extrapolating from the evidence in our study, we speculate that ADM limits the formation of scar tissue and reduces esophageal stricture rate, possibly via diminished submucosal fibrosis, enhanced re-epithelization, and attenuated inflammatory cell/mediator infiltration.

This is the first study to evaluate the efficacy and safety of FCSEMS with ADM for the prevention of esophageal stricture. We conducted a robust post-endoscopic stricture validation with full circumferential wounds in a porcine model, while most studies have utilized a semi-circumferential ESD or smaller wounds. Moreover, ADM and FCSEMS were stably consolidated, and ADM could be placed at the wound site with the stent intraoperatively, where no metal clip or fibrin glue was required to secure the ADM graft attachment, which greatly decreased the complexity of ADM application. Simultaneously, we carried out a longer observation period (4 weeks) for acquiring more data.

While ESD is a common treatment option for early esophageal cancer, post-procedure stenosis has remained as its primary limitation. Moreover, the mechanisms underlying the development of an esophageal stricture post ESD remain to be comprehensively elucidated. Typically, the repair of post-ESD esophageal mucosal defect involves three phases: inflammatory response, epithelial proliferation, and extracellular matrix remodeling (Broughton et al., 2006). Honda et al. (2010) performed an animal study, from which they speculated that fibrosis formation occurs in the submucosal and muscularis propria during the mucosal defect healing process, which results in the decrease of the esophageal wall elasticity and movement. Consequently, esophageal stenosis is formed following this cascade. Thus, an effective prophylactic maneuver for an esophageal stricture is aimed at maintaining or re-establishing wound equilibrium and allowing the satisfactory healing process to occur by inhibiting the inflammatory response, promoting epithelial regeneration, and preventing muscularis propria injury.

Currently, tissue engineering and regenerative medicine approaches are targeted toward the reconstruction of structurally and functionally normal tissues, including tissue shielding methods with polyglycolic acid (PGA) sheets and fibrin glue (Sakaguchi et al., 2016), or autologous tissue transplantation (Chai et al., 2019). PGA sheet, a biodegradable suture material, can provide plentiful cytoskeletons that help in drug loading and cell crawling during the repair process, thereby leading to a decreased risk of scar germination and stricture formation (Chai et al., 2018). Nevertheless, these approaches require laborious methods to prepare shielding materials, and are associated with an unstable and complex technology.

ADM is a soft tissue graft created by the decellularization of tissue, which leaves the extracellular matrix in its original undamaged state while removing all materials which could potentially trigger a host immune response (Cornwell et al., 2009). In the current study, the ADM was available from bovine dermis, containing predominantly type I collagen. Additionally, it is produced under aseptic conditions, ensuring that the allograft is sterile. The ADM has been used successfully for wound healing and tissue repair, as well as tissue reconstruction (Jansen et al., 2013; Kirsner et al., 2015). However, few studies have focused on the repair of the gastrointestinal mucosal layer, particularly that of the esophagus. In our preliminary study (Han et al., 2017), we demonstrated the safety and efficacy of placing ADM with metal clips in a porcine model to prevent esophageal stricture formation after a semi-circumferential ESD; however, it was difficult to fix the ADM patch with metal clips on the circumferential wound.

Maintaining the organ morphology is of great significance for organ regeneration in vivo. To facilitate wound healing, the esophagus must be maintained in a tubular form for a certain period. Thus, the current study emphasizes how the ADM patch can be grafted on full circumferential wounds stably and easily, and how the esophageal tubular structure can be maintained. Therefore, we developed a novel stent, the ADM-FCSEMS, which provides the advantage of extraordinary water solubility of the ADM, which allows the adherence to moist wound surfaces with great rapidity. Therefore, no clip or fibrin glue is needed to secure sheet attachment. Additionally, ADM increases friction and adhesion between the stent and wound against stent migration, and the FCSEMS provides a radial force, which fixes the ADM in a stable manner and maintains the esophagus in a tubular form with minimal deformities.

The study was designed and conducted with careful attention to every process and material used; however, there were still several limitations. First, cases of mucosal defects formed after circumferential ESD with 2–3 cm lengths are uncommon in actual clinical settings. Considering the limited esophageal length in miniature pigs and our successful validation of stricture formation, we decided to resect 1-cm-long mucosa in the study. A mucosal defect of approximately 2-cm was formed due to esophageal ductility. Second, the consideration of weight variation as an evaluation outcome was imprecise, and despite development of obvious dysphagia at week 4 in Group B, weight loss in both groups was statistically insignificant. In addition, there was still one case of stenosis in group A at week 4, which may be related to the lack of new epithelium with sufficient thickness. Hence, our following study would focus on whether 2 weeks of stenting is appropriate and whether a longer period of stenting is needed.

In conclusion, FCSEMS with ADM effectively prevented post-ESD esophageal stenosis possibly by the following mechanisms: diminished submucosal fibrosis, enhanced re-epithelization, and attenuated inflammatory cell/mediator infiltration at the microscopic level. We estimate that our results could provide important preclinical experience for subsequent clinical trials. Planning is underway for a first-in-human study to assess the feasibility of ADM-FCSEMS placement for the prevention of post-ESD esophageal stricture in patients with early esophageal cancer.