Adipose-derived Mesenchymal Stem Cells Improve the Healing of Tracheoesophageal Fistula in a Beagle Model

Background Tracheoesophageal stula (TEF) is still a devastating clinical problem with high mortality. New clinical strategies were developed to sustain survival time. Mesenchymal stem cells were applied in many clinical wound healing elds. This study aims to investigate the main effect of stem cell to TEF. Material and Method We established a beagle model with TEF by punching the trachea and esophagus membrane and suturing. The beagles were divided into three groups (group 1 = 1, group 2 = 6, group 3 = 6). Group 2 and 3 received a TEF building operation. Group 3 were injected 2 ml stem cells (10 6 per animal), and group 2 injected saline water with same volume. Group 1 did not receive any intervention. All animals receipted total parental nutrition. The closure degree of stula tissue was observed by bronchoscope and post-mortem after 35 d. Result Morphologic and histopathologic changes of stulas were assessed by gross and endoscopic observation. The stulas diameter was measured. In 35 d postoperatively, group 2 showed that 3 animals died for acute smother, 2 animal died for severe chronic pneumonia and 1 animal with consistent stula. Group 3 showed that 2 animals fully closed, 3 animals stula diameter signicantly decreased, and 1 animal died for acute smother. In autopsy result, group 2 animals showed severer pneumonia degree than group 3 animals in 35 d. Conclusion The transplantation of stem cells can promote healing degree of TEF without any complications and relieve pneumonia at the same time.


Introduce 1. Tracheoesophageal stula
Tracheoesophageal stula (TEF) is a disabling and devastating clinical problem with high mortality in pulmonary and critical care medicine. The majority of stulas origin from malignant diseases including tracheal cancer and esophageal cancer [1] . The injury by tracheal intubation after mechanical ventilation is another reason of iatrogenic TEF [2] . Despite various therapeutic efforts like surgery, endoscopic, and radiotherapeutic strategies, the median survival time is 35 d according to a past research [1] . The patients may die within 3~4 months without treatment timely [3][4][5] .

Main clinical strategies
Recently, the clinical strategies include kinds of surgeries and endoscopic interventions, the patients are still confronted with poor prognosis. Endoscopic strategies are useful in small stula (<1cm) with mild infection by using plastic endoprostheses or cuffed plastic stents [3,6,7] , and the stula (>1cm) with severe local infection should be surgeon according to precious studies. Hence, surgery is the common clinical strategy for current TEF. However, traditional surgical procedure may cause large drainage wound, severe pain, and slow healing, and complicated operative procedure leads to a high rate of postoperative complications and mortality [8,9] . Other alternative operation needs the autologous aps and other material with adequate blood supply [4,5,10] . The availability of pedicled ap is limited by the guaranteed blood supply. Therefore, the TEF need the breakthrough in clinical strategies. Among the new cures, stem cell transplantation is a research highlight.

Material And Method
Animal Six male beagles 3~4 years old, 8~11 kg were obtained from Gangdong laboratory animals monitoring institute. These animals were divided into three groups: group 1 (a blank group, n=1 each, normal animal), group 2 (a control group, n=6 each, TEF operation and saline water injection), group 3 (an experimental group, n=6 each, TEF operation and stem cell injection). These animals were randomly divided into three groups and numbered (group 1: 1101, group 2: 2101-2106 and group 3: 3101-3106).

Mesenchymal stem Cells
Mesenchymal stem cells were selected adipose-derived mesenchymal stem cells (AMSCs) from Saliai Stem Cell Science and Technology Co.LTD. To con rm the AMSCs maintain their phenotypic characteristic after growth in culture. The MSCs were tested for cell-surface marker via ow cytometry (Beckeman-Coulter, Danvers, MA, https://www.beckmancoulter.com ) and found positive for CD73, CD90, CD105 and negative for CD34, CD45, BL and HLA-DR.

Preoperative preparation
The surgical procedure was performed under general anesthesia using intravenous injection of ketamin (10 mg/kg) and thialazine hydrochloride (1 mg/kg). as soon as the dogs lost consciousness an intravenous line was established. Anesthesia ventilation uses iso urane (1%~5%) maintain anesthesia. All animals were intubated and received supplemental oxygen via the tube (30 ml/kg, 15/min). during the operation the dogs had electrocardiographic monitoring.

TEF model construction
The anesthetized animal was placed on the table. The skin incision was mad parallel to the medial margin of the left sternocleidomastoid muscle until the trachea was reached from a lateral approach. The trachea was resected completely from the cricoid cartilage down to the jugulum, the esophagus was also dissected from the connective tissue. Then the trachea was rotated 90° to the right to expose the membranous wall of trachea. A hole was burned by an electrotome into the membranous part, a corresponding puncture was burned into the anterior wall of the esophagus. The mediocre punch was used to expend and form a stula 8 mm long and 8 mm wide in trachea and esophagus. Then the margin of the esophageal puncture was sutured to the margin of the tracheal stula. This resulted in a communicating stula between trachea and esophagus. The subcutis and cutis were closed with 3.0 Prolene sutures. The all animals received 1.0 g Ampicillin sodium as a prophylactic antibiotic after surgery immediately ( g 1).

Stem cell transplantation
During the operation, all group 3 animals were injected 1 x 10 6 AMSCs 2 ml into the tracheal membrane and esophageal submucosa. All group 2 animals were injected equal saline water instead in the same place.

Postoperative care
For the rst 48 h after surgery, the dogs received analgesia with intramuscular Flunixin meglumine 0.5 g/d.
Each week after operation, the bronchoscope is conducted to ensure that there is not hematocele or secretion in trachea. The stula is observed at same time.
The oral intake was forbidden in group 2 and 3 during the postoperative period. The total parenteral nutrition (TPN) was applied in two group. During this period each animal was monitored rigorously. The symptoms like chocking, vomiting, cough aggravate and even smother should be recorded. The group 1 animal use the normal nutrition. The other aspects of animal care are same in three group

Euthanasia and observational indexes
The animal was euthanized by cutting the femoral artery to avoid the interference of blood congestion in lung at 35 d. the animal died before 35 d should record the death time and analysis the death reasons.
The gross stula tissues were cut for the describe of stula parameter in the following section after euthanasia and accidental death.

Fistula parameter
After harvesting the stula form the animal, the trachea and esophagus tissues are opened longitudinally. The diameter of stula was measured by vernier caliper. The change of the diameter was assessed in three groups.

Histology
Tissue sections of stula were immediately dissected using a scalpel into pieces and incubation in 4% paraformaldehyde, embedded in para n, serially sectioned and stained with Masson dye for light microscope.
Statistical analysis SPSS statistics software v25.0 was used for data analysis. Quantitative data are expressed as the mean ± standard deviation (SD).

Symptom, nutrition and survival time
The operation procedure was performed successfully in all 12 beagles. In postoperative period, each TEF animal present symptoms like cough, vomiting and tracheal chocking which induced by the re ux of gastric content especially.
The survive of animal was present in the gure 1. Three control group animals and one experimental animal (2101, 2103, 2106, 3102) were died in the rst postoperative week ( g 1). The stulas in animal died in rst week were not formed according to the autopsy because the TEF in these animals was maintained by the surgical suture rather than brotic stula wall between trachea and esophagus (Fig 2). During the rest experimental period, two animals of control group were died for the pulmonary in ammation, and rest animals in experimental group were survived. These animals (control 1 and experimental 5) were euthanized by cutting the femoral artery in 35 d.

Endoscope
Each week after surgery, the bronchoscope was performed to con rming the location and healing degree of tracheal stula. (picture 3) On the aspect of secretion, excess white foamy secretion was covered the stula and diffused in whole trachea during the early postoperative stage (1~2 week) in all groups.
The difference occurred at the 14 d postoperatively. In the control group, the viscous secretion appeared in the upper trachea and the accumulation of sputum in low lobes was observed during the later postoperative stage (3~5 week). on the contrary, in the experimental group, the volume of secretion and sputum was decreased in the 3 week after surgery, and sustain a condition with few re ux and secretion during the 4~5 week postoperatively (table 1).
The healing degree of stula was consistent with the volume of secretion. In the control group, there is little change about stula diameter under bronchoscope. In the early postoperative stage (1~2 week), the margin stula appeared amount of necrotic tissue in both groups. In the 3 week, the diameter of stula in group 3 was decreased gradually especially in the animal 3104 and 3105 (table 2). On the contrary, the stula in group 2 was still unchanged. This diversity was sustained during the rest period of experimental. The difference was con rmed by the diameter measured from the stula tissue after autopsy.
In pneumonia, the open stula leads to the contamination in lower lobs in animals from control group (table 4). This result also can be con rmed by the follow-up autopsy result. The lung in animal 2102 present a severe in ammation in almost whole lung, and the lung in animal 3104 was in a healthy condition ( g 3).

Histology
As shown in the g 4, the stula models were successfully established with fused constructions, as con rmed by Masson staining and gross examination of stula tissue.
The diameters of each stula were presented in gure 1. In group 3, two stulas were closure (3104, 3105), the rest stulas show the tendency of decreasing the diameter. In group 2, there is little change in stula diameter.
In open stula, the border between trachea and esophageal was well fused together. The border of epithelium was formed in the open stula in the group 2 and 3. It was not founded in the closed stulas in group 3 ( g 4).
There was little secretion in trachea in group 3 compared with group 2 ( g 5).

Discussion
In this study, we aimed to investigate the morphological changes of repairing stula with transplanting AMSCs into stula tissue in an animal model of TEF. Results showed that the AMSCs group received more bene t than control group in healing and in ammation aspects. The use of AMSCs led to less mortality, less rate of pulmonary infection and better degree of closure.
Mesenchymal stem cells (MSCs), which are referred to as stromal progenitor cells are self-renewing and expandable stem cells. MSCs from amount sites, including bone marrow and adipose tissue, may have immunomodulatory properties, thereby indicating a possible use for MSCs in regenerative medicine and tissue engineering [11] .
In contrast to bone marrow-derived stem cells, adipose-derived stem cells are abundant in adipose tissue and are easy to harvest and expand in vitro, which makes them a describe cell type for regenerative cell therapy [12] .
In our result, all animals present severe symptoms like acute cough, airway chocking, and vomiting in the rst postoperative week. The mortality caused by re uxing of gastric contents was different in two group ( g 1). There are not any researches have been reported yet. Combining the subsequent results about pulmonary infection. we suspect the stem cell may reduce the risk of postoperative and re ux mortality in the rst week.
There were many reports that MSCs have been used to cure, including bone repair, cardiovascular disease, plastic surgery, idiopathic pulmonary brosis, chronic liver injury, acute renal injury, and so on [13] .
However, few studies using in vivo animal models had provided evidence on healing effects of MSCs in TEF. There are several researches about the e cacy of various approaches on TEF animal model ( g 2).
In term of survival time, the surgery retains a longest survival rate and time.in Yang et al study [14] , all TEF animals received a long survival time up to 56 d. it is longer than survival time of stent and stem cell. In the research of Wagner [15] , the healing effects of covered stent, the longest survival time was 36 d which still need an additional tracheal stent. The longest survival time by one esophageal stent was only 30 d. Our survival day was 35 d in stem cell group, lower than Yang but longer than Wagner. In the rate of pneumonia, the aspiration pneumonia occurred in Wagner and our study after the rst postoperative week compared with the surgery study. In our experiment, the pulmonary infection presents a severe consolidation in even whole lungs (upper lobes and lower lobes) in the control group. On the contrary, the contamination was concentrated in the upper lobes in the experimental group. The current researches about stem cells have showed the relative curative effect on the pneumonia. But we believe the stula closure effect from stem cells is the main motivation to avoid pulmonary infection. The open stula leads to the persistence of re ux of gastric, which lead to the infection in upper lobs in the early postoperative stage (1~2 week). In the middle stage, the closure of stula prevents the erosion in lung caused by the re ux in the experimental group (3 week). This procedure was not occurred in the control group, which lead the severe lesion in the lower lobs in the later stage. The infection which involved in whole lung bring about the death of two animals (2102, 2014) in control group.
Although there are not reports on the healing research of TEF by stem cells, there are many cases for the bronchopleural stula on the stem cell in the healing of the bronchial stumps.
The earliest report of stem cell in treatment of trachea stula is form Agero et al in 2008 [16] . he reported a case that a bronchopleural stula patient was used autologous adipose-derived stem cell suspended in brin glue to close the stula successfully. That is the rst report on the treatment of tracheal stula by autologous cell therapy. Prudencio et al [17] reported that two patients with bronchopleural stula received a successful closure, and no treatment -related adverse reactions recurred in the 3 year follow up.
Francesco et al [18] reported that bone marrow-derived stem cells were effectively closes bronchopleural stula in animal model by extraluminal broblast proliferation and collagenous matrix development, which is similar with our results. In 2015, he reported the succeed in tracheobronchial dehiscence after stem cell infusion [19] . In 2016, Johnathon et al [20] combined the bio-absorbable mesh and autologous stem cell, and the stem cell-seeded matrix graft was used to repair a multiply recurrent postpneumonectomy bronchopleural stula successfully. That is the rst application of combination with bio-material and stem cell in tracheal stula therapy. In the meantime, the di culty spot we met is also that stem cells lack adhesive carrier in stula transplantation. Partial stem cells injected into submembrane were leaked into outside. The leak of cells might be the reason that the animals in group 3 present a different healing degree. According to our previous research, the cardiac septal defect occlude (Amplazer occlude) has an eminent load capacity of various cells in vitro. However, the occlude alone could not have a long-term curative effect to stula in our precious animal study. The di culty in epithelialization and growth of miscellaneous bacteria are the limitation for the new occlude. Hence, the stem cells might ameliorate the problem according to the good epithelialized result from our research. The combination may be another new clinical strategy of the treatment of TEF.
In the current study, we reported the various effects of MSCs transplantation on TEF such as following: (1) reduce the death risk in early stage of TEF; (2) relieve the pneumonia and reduce the accumulation of secretion; (3) accelerate the closure of TEF in some degree; (4) increase in collagen synthesis because of broblast from stem cell.
In conclusion, this was the rst study to use locally transplanted cell therapy for the healing of TEF. AMSCs therapy signi cantly accelerated the healing degree of TEF. That may be a potential therapeutic strategy to acquired TEF if combine the appropriative stent which can load stem cell e ciently. All data, models, and code generated or used during the study appear in the submitted article.