PRP improves the outcomes of autologous skin graft transplantation on the esophagus by promoting angiogenesis and inhibiting fibrosis and inflammation

Animals

Twenty-seven adult male Bama miniature pigs weighing 25–30 kg were randomly divided into three groups. Nine of them were arranged to place fully covered esophageal stents (FCES) alone (control group). Autologous skin-grafting surgery was performed in the remaining 18 pigs. Animals with/without the application of PRP were subdivided into the PRP-ASG group and ASG group, respectively. All animals were fasted two days prior to the procedure and given water ad libitum. Three pigs in each group were euthanized on Days 7, 14, and 28. All experimental procedures were approved by the Institutional Animal Ethics Committee of the First Medical Center of PLA General Hospital.

Preoperative preparation

Light anesthesia was induced using intramuscular injections of xylazine hydrochloride (25 mg/kg) and midazolam (0.25 mg/kg). General anesthesia, introduced with 3–4% sevoflurane and maintained with 0.5–2.0% sevoflurane, was achieved by endotracheal intubation. The animals were placed in the left lateral position, and the right side of the pig dorsum was shaved and then disinfected with 70% ethanol and iodine tincture.

Artificial esophageal defect

A schematic overview of the modified autologous skin-grafting surgery on the esophagus is shown in Scheme 1.

Scheme 1

Schematic diagram of the surgical procedure.

ESTD was performed using an endoscope with a single-accessory channel (GIF-Q260J, Olympus, Tokyo, Japan). A transparent cap (D201-11804; Olympus) is routinely placed on the tip of the endoscope. A mixture of 0.9% saline and 0.1% methylene blue was used for submucosal injection. Various electrosurgical knives, including a triangle-tip knife (KD-640 L; Olympus), a dual knife (KD-650Q; Olympus) and an insulated-tip (IT) knife (KD-611 L; Olympus), were applied to complete the procedure. Bleeding was controlled by the application of hemostatic forceps (FD-410LR; Olympus). All pigs underwent esophageal ESTD 30–35 cm from the incisor. Circular dots were made at 30 cm and 35 cm to reveal the oral and anal margins with electrosurgical knives, respectively. Then, a dual knife was used to separate the mucosa transversely after the submucosal fluid cushion reached a feasible level on the anal end and the oral end in turn. A dual knife was used to form a submucosal tunnel with the assistance of repeated submucosal injections from the oral side to the anal side. Next, lateral mucosal resection close to the markings was completed by an IT knife from the anal to the oral side until a length of 5 cm circumferential esophageal mucosa was removed (Figure 1A).

Figure 1

Artificial circumferential mucosal defect followed by endoscopic injection of PRP and transplantation of ASG. (A) Artificial circumferential mucosal defect on esophagus (5 cm in length). (B) Split-thickness skin graft was harvested from the right back of the pig. (C) The graft was sewn into an “oversleevelike” skin and was covered the outside of esophageal stent. (D) Submucosal injection of PRP. (E) The stent was placed at the location of the esophageal defect. (F) Using endoscopic clips to fix the stent. PRP: plasma-rich platelet; ASG: autologous skin graft.

Skin graft

Using a Humby knife, a 6.5 × 8.0 cm split-thickness skin graft (Figure 1B) was harvested from the right side of the pig dorsum by a plastic surgeon and then sewn on a FCES (EVO-FC-20-25-12-E, Cook Medical, Bloomington, USA) into an “oversleeve-like” skin with absorbable suture (VICRYL Plus; 4–0; 1.5 Ph. Eur) (Figure 1C). A No. 11 scalpel was used to make well-distributed holes in the skin graft at 2 cm intervals.

Preparation and endoscopic injection of PRP

Fifty milliliters of blood was collected from the tail vein of pigs in the PRP-ASG group and then transferred to a test tube containing 0.5 mL heparin (No. 1 Biochemical & Pharmaceutical Co., Ltd., Shanghai, China). Two steps were used to obtain the PRP: centrifugation at 2000 × g for 10 min to obtain a supernatant and recentrifugation at 2350 g for 10 min to remove the upper two-thirds of the plasma and retain the lower one-third as PRP (approximately 6 mL).^[11] Prior to the transplantation of the autologous skin graft, 0.5 mL PRP was slowly injected into the uncovered submucosal layer or muscular layer at 12 regular intervals via the submucosal injection needle (NM-200 L; Olympus) (Figure 1D).

Stent placement

An FCES (with/without skin graft) was deployed in vivo to the location of the esophageal defect (Figure 1E). The specific processes were as follows. Grasping forceps were passed through the biopsy channel of the endoscope to tightly grasp the distal steel lasso loop of the FCES. Then, under direct visualization of the endoscope, the FCES was delivered to the esophageal ulcer. Finally, the middle section of the FCES was fit into the esophageal defect by releasing the steel lasso loop. Furthermore, to prevent stent migration, five or more endoscopic clips (Rocc-D-26-195; Micro-Tech, Nanjing, China) were applied to fix the stent to the esophageal wall (Figure 1F).

Postoperative care and follow-up

All animals were administered intramuscular injections of 80 mg/day gentamicin (Baiyunshan Tianxin Pharmaceutical, Ltd., Guangzhou, China) for 3 days postoperatively. They were all fasted for 1 day and fed a liquid diet for 5 days until they were able to consume a semisolid diet. Prior to euthanasia, all of them underwent a repeat endoscopy each week to define the position of the stent. If stent migration occurred, endoscopic forceps were used to place the stent in the original position, and then, the stent was fixed with clips. The total number of clips used before killing was recorded.

General evaluation

Weight loss was recorded from the body weights measured before endoscopic resection and those before euthanasia. Comparisons were performed between the three groups. Dysphagia was scored by using a 5-point scale before euthanasia: 4 = complete dysphagia, 3 = able to swallow liquids only, 2 = able to swallow a semisolid diet, 1 = able to swallow some solid food, 0 = normal swallowing (able to eat a normal diet).^[12]

Macroscopic assessment

The entire esophagus was removed and dissected longitudinally after the animals were euthanized. To evaluate the survival rate of skin grafts, photographs of the entire esophagus taken in each group were analyzed by ImageJ software (National Institutes of Health, USA). The rate of skin graft survival was calculated by the formula (A₀-A₁) /A₀ × 100%, where A₀ and A₁ were the total wound area and surviving skin graft area, respectively.^[13] The degree of stenosis at the artificial ulcer was expressed as the lateral mucosal contraction rate and was calculated using the following formula {1-[Length of the short at site of maximal contraction/ (Length of short axis at a normal mucosal site on upper side + Length of short axis at a normal mucosal site on lower side] × ½) } × 100%.^[14]

Histological analysis of reconstructed tissues

The esophageal specimens were fixed in 10% formalin for at least 48 h and then dissected routinely to prepare paraffin blocks. Three micrometer-thick sections were used to perform hematoxylin and eosin (H & E) staining with conventional methods. In each sample, 12 locations (× 200 magnification) with regular intervals were randomly viewed by two senior histologists who were blinded to the origin of the sections, and then the total number of inflammatory cells was recorded and compared between the three groups.

Evaluation of angiogenesis

Cluster of differentiation 34 (CD34; GeneTex, USA) and VEGF (Abcam, Cambridge, UK) were detected by immunohistochemistry (IHC). The sections were deparaffinized in xylene 3 times and rehydrated in a graded series of ethanol with decreasing concentrations. We used Tris-buffered saline (TBS; pH 7.4) for rinsing. Then, the cells were blocked with 5% serum for 30 min before incubation with porcine primary antibodies against CD34 and VEGF at 4°C overnight. Tissue slices were further incubated in secondary antibody at 25°C for 50 min, visualized with diaminobenzidine and counterstained with hematoxylin. Twelve random fields (× 200 magnification) of each section were captured by optical microscopy. The integrated optical density (IOD) of VEGF was detected using ImageJ software.^[15] The microvessel density (MVD) was assessed by two histologists counting the number of positive vessels directly from CD34^-stained images.

Evaluation of submucosal tissue fibrogenesis

Expression of alpha smooth muscle actin (α-SMA; Novus Biologicals, USA) was detected by IHC. The images were taken at 200× magnification with 12 regular intervals and then used to measure the IOD. Masson’s trichrome staining was used to measure the thickness of submucosal tissue, detect the percent of collagen content and grade the damage of fibrosis. Submucosal tissue thickness was defined as the distance between the basal cells of the epithelium and the esophageal muscular layer and was calculated from 12 regular intervals in each sample. The collagen volume fraction was identified as the ratio of collagen to the remolded tissue. Images of Masson’s trichrome staining (× 200) were taken by optical microscopy, and the collagen content was semiquantitatively evaluated by analyzing the IOD via ImageJ software.^[16] The damage of fibrosis was graded using the following scale: 3 = transmural fibrosis of the muscularis propria, 2 = atrophy or fibrosis present but confined to the outer longitudinal muscle layer, 1 = atrophy or fibrosis present but confined to the inner circular muscle layer, and 0 = no atrophic or fibrotic change in the muscularis propria evident in any of the examined sections.^[17] Collagen I (COL-1) and collagen III (COL-3) fibers were shown by picrosirius red connective tissue staining. Under polarized light, COL-1 was identified by its thick fibers and compact and dense fibrils, which were yellow to light red colored. COL-3 formed green fine fibers with thin fibrils.^[18] These sections were visualized using a microscope digital camera, and the ratio of COL-1/ COL-3 was detected using Image-Pro Plus software (Media Cybernetics, USA).

Statistical analysis

Results are described as the means ± standard deviations or median. For comparison of quantitative variables between the different groups, ANOVA with repeated measures was performed. Nonnormal quantitative variables were compared using the Kruskal-Wallis test or chi-square tests. A P value < 0.05 was considered significant. All statistical analyses were performed using SPSS for Windows 26.0 (IBMCorp, Armonk, NY).

Macroscopic appearance and clinical evaluation

Circumferential ESTD was performed safely in each group. Endoscopic observation after stent removal (Figure 2A) and macroscopic appearance (Figure 2B) of the reconstructed esophagus on Day 28 showed the presence of the artificial ulcer and confirmed that the majority of the grafted skin survived. The average survival rate of the grafted skins was significantly higher in the PRP-ASG group than in the ASG group on Days 7, 14, and 28 (98.6%, 94.8%, 92.2% vs. 72.1%, 60.6%, 55.1%, P < 0.01) (Figure 2C). The mucosal contraction rates on Days 7, 14, and 28 were 2.37 ± 0.47%, 5.23 ± 0.47% and 7.83 ± 0.51% in the control group, 2.03 ± 0.63%, 3.00 ± 0.36% and 4.43 ± 0.25% in the ASG group, and 1.70 ± 0.27%, 1.87 ± 0.21% and 2.13 ± 1.10% in the PRP-ASG group, respectively. The PRP-ASG group had a significantly lower mucosal contraction rate on Day 28 than the other two groups (P = 0.013), with no difference on Days 7 and 14 (Figure 2D).

Figure 2

Endoscopic observation and macroscopic appearance of the remolded esophagus on day 28. (A) Typical macroscopic appearance of the esophageal lumen in each group. (B) Endoscopic observation of the esophageal lumen in each group. Statistical analysis of (C) skin graft survival rate, (D) mucosal defect rate, (E) dysphagia score, (F) weight loss, (G) ratio of weight loss, (H) number of endoscopic clips used. ^##P < 0.01, ^###P < 0.001, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001. PRP: plasma-rich platelet; ASG: autologous skin graft.

Although slight swallowing difficulties occurred in each group on Day 28, no pigs developed severe stricture, and the dysphagia score was comparable between the different groups (Figure 2E). Weight loss was significantly higher in the control group than in the ASG group and the PRP-ASG group on Days 7, 14, and 28 (0.97 ± 0.35 kg, 1.37 ± 0.21 kg, 2.57 ± 0.25 kg vs. 0.33 ± 0.12 kg, 0.57 ± 0.15 kg, 1.83 ± 0.31 kg vs. 0.27 ± 0.12 kg, 0.47 ± 0.06 kg, 1.47 ± 0.15 kg, P < 0.05) (Figure 2F). The ratio of weight loss was in agreement with the weight loss (Figure 2G). There was no significant difference in the number of endoscopic clips used among the three groups (Figure 2H).

Histological findings

Histological analysis of recombinant mucosa showed that a stratified and mature epithelium closely resembled the surface of the esophagus, indicating the good growth and histocompatibility of the skin graft (Figure 3A). The inflammatory cell count demonstrated that the PRP-ASG group had a considerably lower accumulation of inflammatory cells than the other two groups on Days 7, 14, and 28 (Figure 3B). As shown in Figure 3C, massive inflammatory cell infiltration was observed in the remaining submucosal layer in the control group. The number of inflammatory cells gradually decreased with time, and leukocyte infiltration was always severe in the control group, moderate in the ASG group, and mild in the PRP-ASG group.

Figure 3

Histopathological finding of the central artificial ulcer sites. (A) Magnified image of the recombinant mucosa. (B) The count of inflammatory cells. ^*P < 0.05, ^****P < 0.0001. (C) Representative images of the central endoscopic resection sites. Scale bar = 100 μm.

Angiogenesis evaluation

As shown in Figure 4, the number of CD34^-positive microvessels and the proportion of VEGF-positive cells in the submucosal tissue on Days 7 and 14 were high in the PRP-ASG group, whereas the number of CD34^-positive microvessels and the proportion of VEGF-positive cells in the ASG group and control group were relatively small. On Day 28, the MVD and expression level were similar to those of the ASG group but lower than those of the control group. Overall, the PRP-ASG group had stronger angiogenesis in the reconstructed tissue within 14 days.

Figure 4

Angiogenesis-related immunohistochemistry staining and histological evaluation on day 7, 14, and 28. (A) CD34-stained sections of the three groups and (B) quantification of microvessel density. (C) VEGF-stained sections of the three groups and (D) quantification of IOD. ^****P < 0.0001. Scale bar = 100 μm.

Fibrogenesis assessment

On Masson’s trichrome staining, the wound tissue maturity evaluated by the degree of collagen volume fraction demonstrated that submucosal tissue repair in the PRP-ASG group had a significant advantage compared with the other two groups within 28 days (Figure 5A and B). Picrosirius red staining demonstrated that the ratio of COL-1/COL-3 on Day 28 in the PRP-ASG group was significantly lower than that in the other two groups, and the ratio was in agreement with the ratio of the normal tissue (Figure 5C). The results revealed the presence of large amounts of COL-3 deposition and small amounts of COL-1 deposition in the PRP-ASG group (Figure 5D).

Figure 5

Measurement of collagen volume fraction and collagen birefringence under polarized light microscopy. (A) Representative images of Masson’s trichrome staining in each group and (B) quantification of collagen volume fraction. (C) Analysis of ratio of COL-1 to COL-3 and (D) different types of COL indicated by Picro-sirius red staining. Scale bar = 100 μm. ^****P < 0.0001.

As shown in Figure 6A, the number and proportion of α-SMA-positive cells were low in the PRP-ASG group, moderate in the ASG group, and high in the control group. Statistical analysis of the IOD of α-SMA indicated that the cell ratio of α-SMA-positive cells to total spindle mesenchymal cells in the PRP-ASG group was significantly lower than that in the ASG group and control group (Figure 6B). At the muscle layer, fibrotic tissue invaded through the inner muscularis in the control group, whereas fibrosis was mild in the two transplanted groups. The damage score for the muscularis propria on Day 28 in the control group was significantly higher than that in the ASG group or PRP-ASG group (3 vs. 0, P < 0.05) (Figure 6C). Perpendicularly, the thickness of the submucosal tissue on Day 28 in the PRP-ASG group was significantly thinner than that in the ASG group and control group (1.24 ± 0.12 μm vs. 1.78 ± 0.11 μm vs. 1.96 ± 0.09 μm, P < 0.0001) (Figure 6D and E).