TY - JOUR
AU - PhD, Stuart Williams,
AB - Background: Anorectal fistulas (ARFs) are a common, devastating, event in the life of a patient with Crohn's disease. ARFs occur in up to 50% of patients with Crohn's disease. Treatment begins with surgical drainage of the initial abscess, followed by antibiotic therapy, then anti-inflammatory medications. If medical therapy fails to close the fistula tract, surgical intervention is often pursued. Surgery incurs risk of incontinence because of sphincter injury. Increasingly, the role of cell-based therapy is being investigated in ARFs. We evaluated the role a bioabsorbable scaffold plays in delivering cell-based therapy using a porcine model of AFR. Methods: ARFs were mechanically created and matured by setons. After 28 days, setons were removed; periaortic fat was harvested and processed for stromal vascular fraction (SVF). The cells were labeled with a membrane stain for later identification, then injected into the fistula or implanted through scaffold. Fistulas not treated with cells were injected with fibrin glue. Animals were monitored visually for healing at weeks 2 and 4, then euthanized to evaluate fistulas for histologic healing. Results: All fistulas (6/6) treated with SVF + scaffolds healed by week 2, compared with only 4/6 with just SVF and 0/5 treated with fibrin glue. Scaffolds retained SVF within the fistula tract more readily than injection method and SVF+scaffold treatment accelerated the healing process. Robust neovascularization was also seen in fistulas treated with SVF+scaffold. No adverse events occurred. Conclusions: Scaffold technology may improve cell-based therapy healing rates for Crohn's ARFs. This advance should be investigated by human trials. Crohn's disease, anorectal fistula, stromal vascular fraction, cell-based therapy The onset of anorectal fistulas (ARFs) in a patient with Crohn's disease (CD) marks a critical turning point in their disease course. Most patients with CD initially present with inflammatory manifestations of the gastrointestinal tract, but over time uncontrolled inflammation leads to structural changes in the bowel that can result in a breach of the bowel wall and ensuing sepsis.1,2 This process leads to the development of an abscess outside the bowel wall. In the region of the rectum, the drainage tract often involves the anal sphincter. Although tissue damage resulting from the primary inflammatory process alone can cause sphincter injury, surgical treatments frequently compound that injury, leading to impaired anal continence.3 Unfortunately, medical treatments fail to heal more than 33% to 50% of active fistulas4,5 and even complex surgical procedures such as advancement flaps6 or ileostomy7 have limited long-term healing rates, complicated by high relapse rates within 6 months.8,9 Clearly, patients affected by ARF need more effective and safe alternatives to supplement currently available treatment regimens. Animal models of fistulizing CD, such as the SAMP1/YitFc mouse model10 or anal furunculosis in canines,11 generate fistulas at random. Size considerations make the murine model more suitable for testing medical therapies than surgical interventions. A recent article describing the treatment of spontaneous canine fistulas with expanded human embryonic stem cells supports the concept that canine models could be used for testing cell-based therapy, but in the absence of a predictable time of onset or inability to readily induce fistulas, this model has limited applicability to preclinical research.12 Based on the limitations of current animal models, we used a modified version of the mechanically induced porcine model of ARF to investigate a new method for treating ARF with cell-based therapy. Adipose-derived mesenchymal stem cell (AD-MSC) implantation for treating ARF has been evaluated in several clinical trials.13–16 AD-MSC–based interventions begin with the expansion of a population of plastic adherent cells isolated from fat that exhibit 2 important stem cell-like properties: stromogenesis and immune modulation.17 Interest in this cell population evolved from case reports describing rapid healing of graft-versus-host disease–associated mucosal lesions after MSC infusions.18,19 A subsequent retrospective study of patients receiving MSC treatment for hematologic malignancies, who were also coaffected by inflammatory bowel disease, reported complete relief of the subjects' inflammatory bowel disease symptoms.18 After this observation, several prospective studies of MSC therapy for inflammatory bowel disease were reported, documenting complete and long-lasting remission of medication refractory CD.20–22 Since then, enthusiasm for MSC treatment for luminal CD has extended to ARF (both those associated with CD, as well as, cryptoglandular fistulas). After the report of a single case of fistulizing CD successfully treated by MSC transplantation,23 subsequent manuscripts have expanded the evidence supporting this therapy.24,25 The promise of long-term closure rates with minimal side effects makes MSC for fistulizing CD an attractive alternative to currently available conventional therapy. However, shortcomings of the AD-MSC based approach, such as the prolonged interval between harvest and implantation, plus the need for expansion facilities operated under current Good Manufacturing Principles and other regulatory guidance, reduce their availability to the average patient. To this end, we conducted a pilot evaluation of a primary cell isolate from known as stromal vascular fraction (SVF) for the treatment of ARFs, using a modified porcine fistula model.26 The use of a primary cell isolate avoids the significant delay between harvest and implantation associated with culture expanded AD-MSC. Using emerging point-of-care technology for processing, this stem-cell rich component of abdominal wall white adipose tissue (WAT) can be harvested just before the implantation procedure.27 In the future, this may translate into a scenario where a clinical evaluation of the fistula tract during a simple examination under anesthesia can rapidly transition to treatment with the administration of freshly harvested SVF. To date, treatment protocols using cell-based therapy for ARF have used a direct tissue injection or infusion technique, which may not be the optimal method for cell delivery. Therefore, we aimed to test a novel method for SVF delivery, using a scaffold as a cell delivery vehicle, to treat mechanically created ARFs in a modified porcine fistula model. Hypothesis We postulated that fistula healing using freshly harvested SVF could be improved by using a novel bioabsorbable scaffold to support cell delivery and retention. Materials and Methods Methodology We designed a pilot trial to investigate how the method of delivering SVF to an induced fistula would alter the treatment outcomes. To test this question, we modified a validated and clinically relevant technique for creating ARFs, which was previously described by Buchanan et al.26 Using an Institutional Animal Care and Use Committee (IACUC)-approved protocol, we studied a total of six6 healthy adult female domestic Hampshire cross pigs (55–65 kg) to compare 3 treatment regimens for the induced ARFs. We mechanically induced fistulas as detailed below. Because these mechanically induced fistulas can heal spontaneously after 4 to 6 weeks, we elected to analyze the primary outcome early in the healing process, to detect any difference between the treatment protocols. The primary endpoint was chosen to be healing at week 2 by visual inspection, whereas secondary endpoints evaluated healing at week 4, histologic healing at week 4, and adverse outcomes from any of the treatment regimens. Fistula Formation Animals were anesthetized, intubated, and anesthesia maintained with isoflurane administered through endotracheal tube for the duration of the following procedures. Fistula tracts were initiated by placing a minimal incision into the ischioanal fossa at the 3, 6, and 9 o'clock positions approximately 2 cm from the anal margin (Fig. 1A), as described by Buchanan et al.26 Although the original model required at least 30 days for spontaneous healing to occur,26 we elected to create a larger, more aggressive fistula that more closely resembled clinical fistulas. We modified the original technique by directing a 14 G intravenous catheter through skin incisions at 3, 6, and 9 o'clock positions on the perineum. After crossing the muscles of the anal sphincter muscle (as confirmed by palpation), the angle was adjusted to ensure that the catheter exited the colon wall proximal to the dentate line. Once the catheter penetrated the rectal mucosa, the introducer was removed and a guidewire was placed through the lumen of the catheter. Next, a 20-F dilator was passed through the sphincter over the guidewire (Fig. 1B). A large caliber seton (4 mm solid-core silicone cord) was carried by the dilator through the resulting tract. After a visual check confirmed that the dilator was pulled through sphincter muscle, the seton cord was threaded into position. The ends of the cord were trimmed and joined with medical-grade cyanoacrylate glue (Fig. 1C). The setons remained in position for 4 weeks to support the formation of a mature fistula tract. For animal comfort, fentanyl patches were placed on the animal's shaved skin along the flank before the surgical procedure and remained in place for the first 24 to 48 hours. Any animal demonstrating evidence of discomfort also received 0.1 mg/kg buprenorphine I.M. q 8 to 12 hours during the first 24 hours. The animals were cared for by veterinary staff during the fistula maturation stage. Animals returned to the OR on day 28 for seton removal, fat harvest for SVF isolation, and fistula treatment. Figure 1. View largeDownload slide Fistula creation and treatment. A, After a small incision is made at the 3, 6, and 9 o'clock positions of the perineum, a 14 G intravenous catheter is passed from the outside mark, across the anal sphincter complex, to a point inside the dentate line. A guidewire is then threaded through the catheter lumen. B, Next, a 14 F dilator traverses the sphincter complex, creating the mechanical disruption responsible for initiating the fistula tract. C, Four-millimeter silicon cord is passed behind the dilator through the tract and is secured in place for 30 days. D, Removal of the seton reveals a well-developed tract crossing the anal sphincter. E, Periaortic adipose tissue is removed to provide a source of fat for cell isolation. F, After fat is processed by the laboratory method, the isolated cells are injected directly into the wall of the fistula tract. Figure 1. View largeDownload slide Fistula creation and treatment. A, After a small incision is made at the 3, 6, and 9 o'clock positions of the perineum, a 14 G intravenous catheter is passed from the outside mark, across the anal sphincter complex, to a point inside the dentate line. A guidewire is then threaded through the catheter lumen. B, Next, a 14 F dilator traverses the sphincter complex, creating the mechanical disruption responsible for initiating the fistula tract. C, Four-millimeter silicon cord is passed behind the dilator through the tract and is secured in place for 30 days. D, Removal of the seton reveals a well-developed tract crossing the anal sphincter. E, Periaortic adipose tissue is removed to provide a source of fat for cell isolation. F, After fat is processed by the laboratory method, the isolated cells are injected directly into the wall of the fistula tract. SVF Harvest Periaortic WAT was used for this project, as this type of pig fat most closely resembles the consistency of human subcutaneous fat, which would be used for treatment of human fistulas. Pig subcutaneous fat is extremely fibrous and is not amenable to the mechanical disruption technique responsible for the removal of SVF. WAT was harvested from the periaortic region immediately after removal of the setons (Fig. 1D). To reach the aorta, the skin and subcutaneous fat were incised under sterile conditions. Next, the muscle layers in the flank region were divided to reach the peritoneal reflection. Taking care not to enter the abdominal cavity, WAT was bluntly dissected from the peritoneal membrane (Fig. 1E). A total of 35 mL of fat was collected from each animal in sterile conical tubes containing ice-cold phosphate-buffered sodium. Flank incisions were closed in 4 layers while the WAT was sent to the laboratory for processing. SVF was immediately isolated using established laboratory procedures.28 Briefly, adipose tissue was washed with Dulbecco's cation-free phosphate-buffered saline, minced and digested with digestive enzyme blends, and diluted with an equal part human serum albumin. The cell suspension was digested for 30 minutes at 37°C with vigorous shaking. Adipocytes were removed by centrifugation (Fig. 2). The AD-MSC–enriched pellet was removed, washed twice in Dulbecco's cation-free phosphate-buffered saline + 0.1% human serum albumin, and resuspended in a sterile Ringer-lactate solution containing 1% bovine albumin and 1 × 107 SVF cells per mL. A portion of the SVF was collected from each animal and assessed for cell yield, percent viability, and flow cytometry. SVF cells were then labeled with DiD intra-vital stain (CellBrite; Biotium, Inc., Fremont, CA) per manufacturer's protocol to allow for identification of the implanted stem cells at the time of histological analysis after fistula tract excision. This stain does not impact viability or function of the SVF cells and is a permanent stain that can be visualized for at least 4 weeks. After final sample preparations were completed within 60 to 90 minutes after harvest, aliquots of the cell suspension containing a 2 × 107-mixed stem cell population (SVF) were administered to the assigned fistula tracts. Figure 2. View largeDownload slide SVF component isolation product and scaffold preparation. The SVF layer is found in the top layer of cells in the conical tube containing processed adipose tissue. Next, the SVF cells are incorporated into the lumen of the scaffold. Finally, the ends were pinched closed with a hemostat (not shown). The scaffold retains water within its lumen once it was sealed, giving the loaded scaffold a “sausage” appearance. The scaffold material handled easily and does not tear during positioning within the fistula tract. Figure 2. View largeDownload slide SVF component isolation product and scaffold preparation. The SVF layer is found in the top layer of cells in the conical tube containing processed adipose tissue. Next, the SVF cells are incorporated into the lumen of the scaffold. Finally, the ends were pinched closed with a hemostat (not shown). The scaffold retains water within its lumen once it was sealed, giving the loaded scaffold a “sausage” appearance. The scaffold material handled easily and does not tear during positioning within the fistula tract. SVF Analysis Cell yield and viability were assessed for each extraction procedure using a NucleoCounter NC-200 automated cell counter (Chemometec, Allerod, Denmark) using a duofluorescent detection method. A small aliquot of each primary isolate was used to perform cell population characterization, assessed with our multicolor flow algorithm using a standard Mesenchymal Stromal Cell Marker kit (Abcam, Cambridge, MA) to evaluate for the following cell populations: CD45−, CD14−, CD44+, CD31+, CD90+, and CD105+. One active population, the AD-MSCs, comprised ∼2% of the cell content of WAT. Flow cytometry was performed on a Becton Dickinson (Franklin Lakes, NJ) LSR II flow cytometer, whereas resulting data were analyzed with FlowJo (Ashland, OR) software. Treatment Delivery Because of the size of the animals treated and operating room restrictions, only 3 animals could be treated at a time. Two groups of animals (n = 3) were treated in back-to-back sessions. In order for each animal to serve as its own control, each treatment regimen had been arbitrarily assigned to a particular position on the perineum, such that each treatment was administered 6 times. Treatment 1 consisted of 2 × 107 SVF cells suspended in 2 cc of sterile Ringers-lactate solution containing 1% bovine albumin directly injected into the walls and lumen of the 3 o'clock fistula (Fig. 1F). The suspension was injected using a long 26 G needle repeatedly inserted to a depth of 2 to 3 mm into the fistula walls of the fistula. We attempted to maintain a consistent depth and pattern of injection by visual and tactile feedback. The first aliquot was injected into the wall of the fistula closest to the internal opening of the fistula, the next aliquot into the walls of the middle portion of the fistula, and the final aliquot into the portion of the fistula closest to the external opening. Treatment 2: the 9 o'clock fistulas received SVF + scaffold treatment, into which 2 × 107 SVF cells were injected into the lumen of an electrospun fibrinogen scaffold. The scaffolds were constructed into sheets from salmon-derived fibrinogen (Sea Run Holdings, Freeport, ME) using an in-house electrospinner under conditions previously described.29,30 After placing the appropriate number of cells on the sheet, the sheet was rolled into a tube. The ends of the scaffold were mechanically sealed by clamping them closed with a small hemostat (Fig. 2). After atraumatic passage of a 5.5-inch–curved Kelly forceps through the lumen of the fistula tract, the scaffold-containing SVF was grasped by the open jaws, then drawn from the inside opening to the outside opening. The inner end of the scaffold was fixed in place within the fistula tract by the placement of a 2/0 Vicryl stitch across the inner opening of the tract, which also loosely approximated the internal opening, but did not seal it closed. A second 2/0 Vicryl stitch was placed across the outer opening and through the outer end of the scaffold, securing the scaffold in place within the fistula tract while maintaining patency of the outer opening. Treatment 3, an active control, involved the injection of 5 mL of commercial fibrin glue prepared according to manufacturer's instructions, with a single stitch used to approximate the internal and external openings. Fibrin glue was used as an active control, as opposed to allowing the fistula spontaneously heal or placing an empty scaffold into the tract, because of concerns that the untreated tract could become infected or that the empty scaffold would become a nidus of infection. Visual Monitoring for Response to Treatment Animals were evaluated for fistula closure by visual inspection on a weekly basis after undergoing fistula treatment. Evidence of closure was determined by visual inspection of the awake pig for epithelialization of the outer opening without accompanying evidence of abscess formation or other accumulation of fecal matter under the epithelial cap. Each animal was evaluated at weeks 2 and 4. Fistula Harvest After 4 weeks, the fistula tracts were harvested. While under general anesthesia, the external and internal closure sites were marked with a subcutaneous/submucosal injection of endoscopic tattoo ink (SPOT; GI Supply, Camp Hill, PA) to correlate the explant with its original anatomical context, so that fistulas would be correctly identified during histologic analysis. Next, the entire anal canal, along with a 2-cm margin and supporting structures (including distal rectum), was excised en bloc and placed in formalin in preparation for histologic analysis. The animals were subsequently euthanized. Histopathology Formalin-preserved tissue explants were imbedded in paraffin. Blocks were prepared for staining by sectioning 5 μm specimens, which underwent hematoxylin and eosin (H&E) staining per routine. Histological examination of the routine H&E staining revealed the deposition of new muscle fibers and the formation of new blood vessels within the cores of the fistula tracts treated with SVF. These changes were further evaluated using immunohistochemistry staining with antibodies to smooth muscle actin (SMA). This antibody stain improves the visualization of smooth muscle cells in a vessel wall. SMA clones of Alpha Sm-1 (Leica Biosystems, Buffalo Grove, IL) were applied according to manufacturer's instructions to 5-μm sections where the core portion of the fistula tract interfaced with the native tissue. SMA stains myocytes in vascular walls, intestinal muscularis mucosae and muscularis propria, and in the stroma of various tissues. It is expressed in myofibroblasts. DAPI (Sigma-Aldrich, St. Louis, MO) staining was performed according to manufacturer's instructions, and DiD and DAPI imaging was performed using a Nikon E800 fluorescence microscope equipped with a Spot digital camera to capture images. Statistical Analysis Comparison between treatment groups was performed by analysis of variance initially, followed by Dunnett's correction for multiple comparisons to detect differences between individual treatment groups. The use of each animal as its own control reduces the variability that can be seen when using a separate control animal. Ethical Considerations All procedures conducted in support of this manuscript were performed in accordance with the University of Louisville Institutional Animal Care and Use Committee approved protocol. Results Fistula Closure as Determined by Visual Inspection The primary endpoint consisted of visual evidence of closure at week 2. Because spontaneous healing occurred well after this time point, we chose week 2 as the primary endpoint to most aggressively differentiate healing from a treatment effect as opposed to spontaneous healing. A secondary endpoint consisted of healing by visual inspection at week 4 and histologic healing determined by evaluation of explanted fistulas. A total of 6 tracts were available for treatment with SVF + scaffold or SVF alone but only 5 for fibrin glue. One animal dislodged a seton from a tract 2 weeks before the treatment date, and the tract was not amenable to treatment. We elected to use the 2 remaining tracts in that animal for cell-based therapy, to maintain the maximal number of tracts available for SVF-based treatments. As documented in Figure 3, visual inspection of fistulas at week 2 demonstrated that 100% of tracts (6/6) treated with SVF + scaffold closed, whereas only 66% of tracts (4/6) treated with SVF alone healed at week 2. When compared with the fact that no fibrin glue–treated tracts healed (0/5), only the SVF + scaffold group healed at a statistically significant faster rate (P < 0.05). By week 4, all SVF + scaffold treated tracts remained closed, whereas all tracts treated with SVF alone had caught up to demonstrate visual evidence of healing, compared with only 40% of fibrin glue tracts (no significant differences). Interestingly, the visual appearance of healing did not predict histologic healing in the case of fibrin glue–treated tracts. Figure 3. View largeDownload slide Visual closure results. Animals were inspected at weeks 2 and 4 for closure of the external fistula opening. All 6 fistulas treated with SVF + scaffold healed according to visual inspection by week 2, compared with 4 of 6 (66.7%) treated with SVF alone, and 0 of 6 in fibrin glue–treated tracts. By week 4, all SVF-treated tracts healed, whereas only 2 of 5 (40.0%) treated with fibrin glue exhibited healing. NS, non significant. *P < 0.05. Figure 3. View largeDownload slide Visual closure results. Animals were inspected at weeks 2 and 4 for closure of the external fistula opening. All 6 fistulas treated with SVF + scaffold healed according to visual inspection by week 2, compared with 4 of 6 (66.7%) treated with SVF alone, and 0 of 6 in fibrin glue–treated tracts. By week 4, all SVF-treated tracts healed, whereas only 2 of 5 (40.0%) treated with fibrin glue exhibited healing. NS, non significant. *P < 0.05. Fistula Closure Determined by Histologic Analysis H&E staining of SVF + scaffold treated tracts revealed a robust healing response in progress, with advanced angiogenesis anchoring the “neo-tissue” to the surrounding native tissue of the fistula tract (Fig. 4). SVF-treated tracts also demonstrated significant progress in healing, but the tissue did not have as mature an appearance as the SVF + scaffold tracts, as they contained a greater number of inflammatory cells than the combined treatment tracts. Immunofluorescence analysis of the 3 treatments revealed important clues underpinning the success of the SVF + scaffold combination over SVF or fibrin glue–treated tracts. Before implantation, SVF cells were labeled with CellBrite Red Cytoplasmic DiD stain, a stable, non-toxic cytoplasmic membrane stain (Biotium, Hayward, CA) that does not transfer from cell to cell. The presence of DiD-stained cells within the healed fistula tract suggests that cells delivered during fistula treatment persist for at least 4 weeks and that at least a portion of the administered cells remain located within the confines of the scaffold material. Based on the relative abundance of DiD-stained cells persisting within the tissue generated at the SVF + scaffold treatment sites (Fig. 5B), it seems that SVF delivery by scaffold results in a greater density of SVF cells within the lumen of the fistula tract, compared with SVF injected into the walls and tract (Fig. 5A). Interestingly, the scaffold delivery device promoted the formation of muscle tissue as evidenced by immature myocytes and vascular outgrowths from the scaffold perimeter that appeared to be integrating into the surrounding tissue (Fig. 4C). Staining with either antibody targeting CD31 (an endothelial cell marker) or antismooth muscle antibody, a smooth muscle cell and myofibroblast marker, demonstrated that the projections most likely corresponded to budding vessels growing from inside the scaffold lumen into the surrounding tissue (Fig. 4C). No indication of residual scaffold could be found, despite the use of birefringence to detect residual crystalline polymer material. The best indicator of scaffold-facilitated tissue regeneration was found in the whorl-like appearance of new matrix and mixed cellular infiltrates. The observation of DiD-stained cells corresponds with retention and integration of transplanted SVF cells within this new matrix. Pockets of immune cells intermingled with smooth muscle cells and stromal tissue and a large number of bright red DiD-stained SVF cells. These cells anchored a whorl-like core of tissue that has formed within the confines of the scaffold wall. The larger number of DAPI staining cells suggests a relative influx of native cells to support the implanted SVF cells. By contrast, fibrin glue–treated tracts histologically demonstrated a lack of tissue regeneration as evidenced by a persistent lumen filled with residual fibrin glue surrounded by an intense inflammatory cell infiltrate (Fig. 5C). The absence of cells by DAPI nuclear staining in Figure 5C provides an additional indication that the fistula lumen persists at least 4 weeks in fibrin glue–treated tracts. Figure 4. View largeDownload slide Tissue and blood vessel development within borders of scaffold. A, The deposition of new smooth muscle, blood vessels, and other stromal components can be seen within the tract of the mechanically induced fistula (black arrows). The orientation of these tissue components contrasts with the orientation of the skeletal muscle native to the external sphincter. B, The inset from A depicts close-up view (×100) of a nascent vessel developing within the stroma filling the tract demonstrates an erythrocyte within its lumen. Rapid assembly of vascular networks from preformed vascular precursors constitutes one of the advantages of treatment with SVF. C, SMA immunostaining highlights both immature myofibroblasts within the stromal tissue, as well as, within the intima of newly formed blood vessels that can be seen penetrating the interface between the scaffold and native tissue (open arrow, ×40). The brown stain results from immune staining, highlighting the cytoplasm of the muscle fiber in the intima of the blood vessels, whereas the blue stain provides a contrast stain of the supporting tissues. Figure 4. View largeDownload slide Tissue and blood vessel development within borders of scaffold. A, The deposition of new smooth muscle, blood vessels, and other stromal components can be seen within the tract of the mechanically induced fistula (black arrows). The orientation of these tissue components contrasts with the orientation of the skeletal muscle native to the external sphincter. B, The inset from A depicts close-up view (×100) of a nascent vessel developing within the stroma filling the tract demonstrates an erythrocyte within its lumen. Rapid assembly of vascular networks from preformed vascular precursors constitutes one of the advantages of treatment with SVF. C, SMA immunostaining highlights both immature myofibroblasts within the stromal tissue, as well as, within the intima of newly formed blood vessels that can be seen penetrating the interface between the scaffold and native tissue (open arrow, ×40). The brown stain results from immune staining, highlighting the cytoplasm of the muscle fiber in the intima of the blood vessels, whereas the blue stain provides a contrast stain of the supporting tissues. Figure 5. View largeDownload slide Histologic features of fistula and SVF retention based on method of implantation. Hematoxylin and eosin (H&E) plus immunofluorescence imaging (×20) of fistula tracts explanted 4 weeks after treatment demonstrates healing pattern and retained labeled cells within the SVF-treated fistula tracts. The presence of DiD-stained cells indicates retained SVF cells. Lipophilic carbocyanine dyes may persist in cultured cells for several weeks and in vivo for up to a year. The dyes do not readily transfer between cells with intact membranes, providing cell migration and tracking information in mixed populations. A, These images demonstrate the effects of SVF injected directly into the walls and lumen. Although the tract has nearly filled, the H&E image documents a small residual lumen. New tissue within the tract appears similar to the images in (B). However, DiD-labeled SVF cells appear in far smaller numbers and aggregate in isolated clumps, rather than achieving a diffuse distribution throughout the tissue as in the scaffold-treated tracts in (B). The interface between disrupted native tissue and newly deposited tissue formed by the cell-based treatment show up most clearly in the H&E images of the SVF alone tract. B, H&E staining of the SVF + scaffold treated tracts demonstrates a whorl-like pattern to the newly generated tissue filling the previous tract. Pockets of immune cells intermingle with smooth muscle cells and stromal tissue and a large number of bright red DiD-stained SVF. These cells are relatively evenly dispersed within the nascent tissue core that has formed within the confines of the scaffold. The large number of DAPI staining cells suggests that other host cells have migrated into the scaffold to support the implanted SVF cells. C, Fibrin glue–treated tracts did not receive any labeled SVF, hence the absence of DiD-labeled cells. However, the residual lumen shows up clearly as a cell-free void in the center of the DAPI and DiD imaging frames. The fibrin glue–treated tracts exhibited a dense inflammatory infiltrate. Figure 5. View largeDownload slide Histologic features of fistula and SVF retention based on method of implantation. Hematoxylin and eosin (H&E) plus immunofluorescence imaging (×20) of fistula tracts explanted 4 weeks after treatment demonstrates healing pattern and retained labeled cells within the SVF-treated fistula tracts. The presence of DiD-stained cells indicates retained SVF cells. Lipophilic carbocyanine dyes may persist in cultured cells for several weeks and in vivo for up to a year. The dyes do not readily transfer between cells with intact membranes, providing cell migration and tracking information in mixed populations. A, These images demonstrate the effects of SVF injected directly into the walls and lumen. Although the tract has nearly filled, the H&E image documents a small residual lumen. New tissue within the tract appears similar to the images in (B). However, DiD-labeled SVF cells appear in far smaller numbers and aggregate in isolated clumps, rather than achieving a diffuse distribution throughout the tissue as in the scaffold-treated tracts in (B). The interface between disrupted native tissue and newly deposited tissue formed by the cell-based treatment show up most clearly in the H&E images of the SVF alone tract. B, H&E staining of the SVF + scaffold treated tracts demonstrates a whorl-like pattern to the newly generated tissue filling the previous tract. Pockets of immune cells intermingle with smooth muscle cells and stromal tissue and a large number of bright red DiD-stained SVF. These cells are relatively evenly dispersed within the nascent tissue core that has formed within the confines of the scaffold. The large number of DAPI staining cells suggests that other host cells have migrated into the scaffold to support the implanted SVF cells. C, Fibrin glue–treated tracts did not receive any labeled SVF, hence the absence of DiD-labeled cells. However, the residual lumen shows up clearly as a cell-free void in the center of the DAPI and DiD imaging frames. The fibrin glue–treated tracts exhibited a dense inflammatory infiltrate. Complications No complications from either fistula creation or treatment were noted. Animals underwent a standard postmortem autopsy to evaluate for significant pathological abnormalities. None were seen with any of the treatment regimens. Discussion This study provides preliminary evidence that SVF delivered by bioabsorbable scaffold promotes faster healing than either SVF alone or fibrin glue. Although the number of animals used in this pilot study was small, the magnitude of the difference between SVF + scaffold and fibrin glue allowed for the detection of a statistically significant result. SVF alone also appeared to heal at a faster rate than fibrin glue, but the difference was not great enough to obtain significance, nor was there a difference between SVF arms at either the 2- or 4-week time points. Although the scaffold could no longer be detected either visually or microscopically, the characteristic whorl-like appearance was seen with each SVF treatment method. The persistence of DiD-stained cells provides another confirmation that we were visualizing the appropriate tissue during microscopic examination. No complications were noted from any of the treatment arms, particularly the SVF + scaffold arm. This is reassuring, as there were some concerns that the presence of the scaffold within the fistula tract could pose a risk as a nidus of infection. Instead, the bulk of the inflammation was seen at week 4 in the fibrin glue–treated tracts (Fig. 5C). The presence of vascular projections suggests that the scaffold delivery mechanism assists in the integration of the tissue “core” into the surrounding tissue. This is particularly relevant given the problem of “ejection” of acellular matrix devices from a fistula tract after a variable period of time.31,32 Interestingly, DAPI staining provided additional evidence to demonstrate that a cell-free lumen persisted in the fibrin glue–treated tracts, despite the presence of visual evidence of healing in 2 of the 5 fibrin glue–treated tracts. This model has several features that limit its applicability to human fistulizing disease, namely the mechanical nature of its origination versus a spontaneous inflammatory etiology for human disease, and the fact that the model fistulas will eventually heal without intervention. Another limitation to the interpretation of the results stems from the fact that we performed a visual inspection of the external site of the fistula treatment in the awake animal. A more extensive evaluation using a probe in a sedated animal would have provided definitive evidence of closure at time points before the actual harvest and tissue analysis. We chose time points that occurred before the onset of spontaneous healing to differentiate SVF-induced healing from spontaneous healing, which rarely occurs in human fistulizing disease, especially those occurring in the setting of CD. Although fistula closure rates in response to medical therapy have demonstrated consistent improvements over time, long-term closure generally requires continuous administration of medical therapy. Surgical therapies, on the other hand, offer the potential for higher temporary closure rates,6–9 but they also have a propensity to injure the anal sphincter and induce incontinence, or require permanent end ileostomy in over 50% of cases.3 Adjuncts to surgical therapy, such as fibrin glue sealants or acellular plugs, have resulted in modest improvements in long-term closure.33–35 Recent reports of cell-based therapy for ARF have documented impressive healing rates, but the technique is resource intensive and requires 2 surgical procedures separated by a prolonged stem cell expansion period. Fistula healing rates range from 71% to 82% at the 8-week time point commonly used as a primary endpoint for this type of trial.14,23,25 The fact that not all subjects respond to cell-based therapies create an unmet need for cell-based therapy that may be improved with innovations in delivery technology. Although our study was a small pilot trial, the results suggest that the combination of a readily accessible fraction of WAT called SVF can be combined with a novel delivery device to accelerate healing rates as defined by the regeneration of tissue. The other benefit of our strategy lies in the fact that autologous SVF can be harvested on the same day as the surgical treatment for the ARF. In this animal model, cell-based therapy alone appeared to more rapidly heal mechanically created fistulas than fibrin glue. Of interest, the fibrin glue stimulated an intense inflammatory infiltrate. Although the sample size of animals tested did not allow for the demonstration of superiority over SVF alone, SVF administered through a novel electrospun bioscaffold clearly healed fistulas more rapidly than fibrin glue. Moreover, SVF retention by the scaffold is likely the reason that imaging with the long-lasting cell membrane stain (DiD) demonstrated a greater SVF density within the fistula than in the case of SVF delivered by the injection technique. The potential for this technology to provide a clinically relevant benefit remains to be proven, but the findings of this investigation suggest potential advantages from the bioscaffold, including the ability to retain cells in the lumen and for those cells to integrate with the surrounding native tissue through capillary budding and vascular integration. Our proposed approach may also greatly improve the clinical feasibility of cell-based therapy for ARF by eliminating the significant delay required for stem cell expansion inherent to other protocols under investigation. SVF also offers other advantages over a pure monoculture of stem cells. Research with SVF from the 1960's demonstrated the presence of a cellular population with similar properties to bone marrow–derived mesenchymal stem cells (BD-MSCs), on the identification of a population of plastic adherent cells capable of multipotent differentiation.17 One key advantage of this particular population, given the name of AD-MSCs, over BD-MSCs, lies in the abundance of these cells: AD-MSCs are found in adipose tissue at levels 100 to 500× higher than multipotent stem cells can found in the bone marrow.17 In contrast to the monoculture of plastic adherent cells expanded to produce either AD- or BD-MSC, the SVF isolated from WAT is a heterogenous mixture of mature adipocyte-depleted cells present in pellet form after processing by mechanical mincing, enzymatic digestion, and centrifugation. The resulting pellet contains both terminally differentiated cells of endothelial and hematopoietic lineage as well as a large number of AD-MSCs (Fig. 6). This mixture of cells possesses a multipotent potential which supports both stromagenesis and immunomodulation.17 These functions are particularly important in healing ARFs, as new tissue needs to be formed within the fistula tract to close the defect, while silencing the highly inflammatory immunological activity of the tissue the lumen of the fistula tract. The immune tone of the local cellular infiltrate is indicated by high levels of proinflammatory cytokines (TNF-α, IL-12) secreted from local immune cell infiltrates.36 A highly beneficial property of the SVF comes from the fact that the composite immunological “tone” of the SVF is generally characterized by an M2 (immunosuppressive and wound healing) phenotype, stemming from a high percentage of M2 macrophages and T regulatory cells.37 Importantly, the AD-MSC isolated along with the SVF possess a large capacity to provide immunosuppressive influences, both in vitro and in vivo.38–40 One salient advantage of SVF over AD-MSC alone or BD-MSC from a wound healing aspect lies in their enhanced angiogenic potential. The AD-MSCs contained in SVF provide more effective restitution in repair models, which has been associated with improved angiogenesis over BD-MSC in a variety of ischemic injury models.41–43 Additional benefits pertinent to fistula healing relate to the rich, heterogenous composite of cells making up SVF, which includes vascular, mesenchymal, and hematopoietic cells. Other important components of SVF pertinent to this project include a population of AD-MSCs comprising 2% to 3% of the total live cells and a CD45− side population enriched with progenitors that possess a strong myogenic potential when introduced into a muscle environment in vivo.44 Both of these populations serve important, complementary roles in the successful healing of ARF, especially in light of the fact that a paucity of migratory myofibroblasts has been associated with nonhealing in CD fistulas.45 Furthermore, exposure to inflammatory cytokines seems to limit the migratory capacity and subsequent contribution to the healing process for this cell type.45 These important details provide support for pursuing additional clinical studies. Figure 6. View largeDownload slide SVF content analysis by flow cytometry. Most cells from SVF, expressing CD45+, represent immune cells isolated from fat. The CD14+ monocyte population skews to an M2, or wound healing, type phenotype, whereas the CD3+ T cell population is balanced between a CD4+ helper T cell phenotype and CD8+ cytotoxic T cell phenotype. Analysis of the CD45−population reveals that 1% to 2% of cells consistently express the traditional CD105+ AD-MSC phenotype. Figure 6. View largeDownload slide SVF content analysis by flow cytometry. Most cells from SVF, expressing CD45+, represent immune cells isolated from fat. The CD14+ monocyte population skews to an M2, or wound healing, type phenotype, whereas the CD3+ T cell population is balanced between a CD4+ helper T cell phenotype and CD8+ cytotoxic T cell phenotype. Analysis of the CD45−population reveals that 1% to 2% of cells consistently express the traditional CD105+ AD-MSC phenotype. References 1. Cosnes J, Gower–Rousseau C, Seksik P Epidemiology and natural history of inflammatory bowel diseases. Gastroenterology . 2011; 140: 1785– 1794 Google Scholar CrossRef Search ADS PubMed  2. Makowiec F, Jehle EC, Starlinger M Clinical course of perianal fistulas in Crohn's disease. Gut . 1995; 37: 696– 701 Google Scholar CrossRef Search ADS PubMed  3. Galandiuk S, Kimberling J, Al-Mishlab TG Perianal Crohn disease: predictors of need for permanent diversion. Ann Surg . 2005; 241: 796– 801 Google Scholar CrossRef Search ADS PubMed  4. Present DH, Rutgeerts P, Targan S Infliximab for the treatment of fistulas in patients with Crohn's disease. N Engl J Med . 1999; 340: 1398– 1405 Google Scholar CrossRef Search ADS PubMed  5. Colombel JF, Schwartz DA, Sandborn WJ Adalimumab for the treatment of fistulas in patients with Crohn's disease. Gut . 2009; 58: 940– 948 Google Scholar CrossRef Search ADS PubMed  6. Jones IT The use of transanal rectal advancement flaps in the management of fistulas involving the anorectum. Dis Colon Rectum . 1987; 30: 919– 923 Google Scholar CrossRef Search ADS PubMed  7. Rehg KL, Sanchez JE, Krieger BR Fecal diversion in perirectal fistulizing Crohn's disease is an underutilized and potentially temporary means of successful treatment. Am Surg . 2009; 75: 715– 718 Google Scholar PubMed  8. Talbot C, Sagar PM, Johnston MJ Infliximab in the surgical management of complex fistulizing anal Crohn's disease. Colorectal Dis . 2005; 7: 164– 168 Google Scholar CrossRef Search ADS PubMed  9. Topstad DR, Panaccione R, Heine JA Combined seton placement, infliximab infusion, and maintenance immunosuppressives improve healing rate in fistulizing anorectal Crohn's disease: a single center experience. Dis Colon Rectum . 2003; 46: 577– 583 Google Scholar CrossRef Search ADS PubMed  10. Rivera-Nieves J, Bamias G, Vidrich A Emergence of perianal fistulizing disease in the SAMP1/YitFc mouse, a spontaneous model of chronic ileitis. Gastroenterology . 2003; 124: 972– 982 Google Scholar CrossRef Search ADS PubMed  11. Houlton J Anal furunculosis: a review of seventy cases. J Small Anim Pract . 1980; 21: 575– 584 Google Scholar CrossRef Search ADS PubMed  12. Ferrer L, Kimbrel EA, Lam A Treatment of perianal fistulas with human embryonic stem cell-derived mesenchymal stem cells: a canine model of human fistulizing Crohn's disease. Regen Med . 2016; 11: 33– 43 Google Scholar CrossRef Search ADS PubMed  13. Ciccocioppo R, Bernardo ME, Sgarella A Autologous bone marrow-derived mesenchymal stromal cells in the treatment of fistulising Crohn's disease. Gut . 2011; 60: 788– 798 Google Scholar CrossRef Search ADS PubMed  14. Herreros MD, Garcia-Arranz M, Guadalajara H Autologous expanded adipose-derived stem cells for the treatment of complex cryptoglandular perianal fistulas: a phase III randomized clinical trial (FATT 1: fistula advanced therapy trial 1) and long-term evaluation. Dis Colon Rectum . 2012; 55: 762– 772 Google Scholar CrossRef Search ADS PubMed  15. Lee WY, Park KJ, Cho YB Autologous adipose tissue-derived stem cells treatment demonstrated favorable and sustainable therapeutic effect for Crohn's fistula. Stem Cells . 2013; 31: 2575– 2581 Google Scholar CrossRef Search ADS PubMed  16. Cho YB, Park KJ, Yoon SN Long-term results of adipose-derived stem cell therapy for the treatment of Crohn's fistula. Stem Cells Transl Med . 2015; 4: 532– 537 Google Scholar CrossRef Search ADS PubMed  17. Casteilla L, Planat-Benard V, Laharrague P Adipose-derived stromal cells: their identity and uses in clinical trials, an update. World J Stem Cells . 2011; 3: 25– 33 Google Scholar CrossRef Search ADS PubMed  18. Le Blanc K, Rasmusson I, Sundberg B Treatment of severe acute graft-versus-host disease with third party haploidentical mesenchymal stem cells. Lancet . 2004; 363: 1439– 1441 Google Scholar CrossRef Search ADS PubMed  19. Ditschkowski M, Einsele H, Schwerdtfeger R Improvement of inflammatory bowel disease after allogeneic stem-cell transplantation. Transplantation . 2003; 75: 1745– 1747 Google Scholar CrossRef Search ADS PubMed  20. Oyama Y, Craig RM, Traynor AE Autologous hematopoietic stem cell transplantation in patients with refractory Crohn's disease. Gastroenterology . 2005; 128: 552– 563 Google Scholar CrossRef Search ADS PubMed  21. Cassinotti A, Annaloro C, Ardizzone S Autologous haematopoietic stem cell transplantation without CD34+ cell selection in refractory Crohn's disease. Gut . 2008; 57: 211– 217 Google Scholar CrossRef Search ADS PubMed  22. Onken J, Gallup D, Hanson J, et al. Successful outpatient treatment of refractory Crohn's disease using adult mesenchymal stem cells. ACG 2006 Final Program Book;2006:121. 23. García-Olmo D, García-Arranz M, Gomez-García L Autologous stem-cell transplantation for treatment of recto-vaginal fistula in perianal Crohn's disease: a new cell-based therapy. Int J Colorectal Dis . 2003; 18: 451– 454 Google Scholar CrossRef Search ADS PubMed  24. Garcia-Olmo D, Herreros D, Pascual M Treatment of enterocutaneous fistula in Crohn's disease with adipose-derived stem cells: a comparison of protocols with and without cell expansion. Int J Colorectal Dis . 2009; 24: 27– 30 Google Scholar CrossRef Search ADS PubMed  25. García-Olmo D, García-Arranz M, Herreros D A phase I clinical trial of the treatment of Crohn's fistula by adipose mesenchymal stem cell transplantation. Dis Colon Rectum . 2005; 48: 1416– 1423 Google Scholar CrossRef Search ADS PubMed  26. Buchanan GN, Sibbons P, Osborn M Experimental model of fistula-in-ano. Dis Colon Rectum . 2005; 48: 353– 356 Google Scholar CrossRef Search ADS PubMed  27. LeBlanc AJ, Nguyen QT, Touroo JS Adipose-derived cell construct stabilizes heart function and increases microvascular perfusion in an established infarct. Stem Cells Transl Med . 2013; 2: 896 Google Scholar CrossRef Search ADS PubMed  28. Aird AL, Nevitt CD, Christian K Adipose-derived stromal vascular fraction cells isolated from old animals exhibit reduced capacity to support the formation of microvascular networks. Exp Gerontol . 2015; 63: 18– 26 Google Scholar CrossRef Search ADS PubMed  29. Boland E, Chang C, Dryden G. Developing a Porcine Fistula Model to Evaluate Cellularized Electrospun Scaffolds. Hilton Head, SC: Regenerative Medicine-Innovations for Clinical Applications; 2011. 30. Boland ED, Matthews JA, Pawlowski KJ Electrospinning collagen and elastin: preliminary vascular tissue engineering. Front Biosci . 2004; 9: 1422– 1432 Google Scholar CrossRef Search ADS PubMed  31. O'Riordan JM, Datta I, Johnston C A systematic review of the anal fistula plug for patients with Crohn's and non-Crohn's related fistula-in-ano. Dis Colon Rectum . 2012; 55: 351– 358 Google Scholar CrossRef Search ADS PubMed  32. Senejoux A, Siproudhis L, Abramowitz L Fistula plug in fistulising ano-perineal Crohn's disease: a randomised controlled trial. J Crohn's Colitis . 2016; 10: 141– 148 Google Scholar CrossRef Search ADS   33. Johnson EK Efficacy of anal fistula plug vs. fibrin glue in closure of anorectal fistulas. Dis Colon Rectum . 2006; 49: 371– 376 Google Scholar CrossRef Search ADS PubMed  34. Champagne BJ, O'Connor LM, Ferguson M Efficacy of anal fistula plug in closure of cryptoglandular fistulas: long-term follow-up. Dis Colon Rectum . 2006; 49: 1817– 1821 Google Scholar CrossRef Search ADS PubMed  35. O'Connor L Efficacy of anal fistula plug in closure of Crohn's anorectal fistulas. Dis Colon Rectum . 2006; 49: 1569– 1573 Google Scholar CrossRef Search ADS PubMed  36. Hadzhiev BD, Murdjeva MA Immunopathological characteristics of cryptoglandular and Crohn's anorectal fistulas. Folia Med . 2010; 52: 5– 12 Google Scholar CrossRef Search ADS   37. Riordan N, Ichim T, Min WP Non-expanded adipose stromal vascular fraction cell therapy for multiple sclerosis. J Transl Med . 2009; 7: 29 Google Scholar CrossRef Search ADS PubMed  38. Yañez R, Lamana ML, García-Castro J Adipose tissue-derived mesenchymal stem cells have in vivo immunosuppressive properties applicable for the control of the graft-versus-host disease. Stem Cells . 2006; 24: 2582– 2591 Google Scholar CrossRef Search ADS PubMed  39. Puissant B, Barreau C, Bourin P Immunomodulatory effect of human adipose tissue-derived adult stem cells: comparison with bone marrow mesenchymal stem cells. Br J Haematol . 2005; 129: 118– 129 Google Scholar CrossRef Search ADS PubMed  40. González MA, Gonzalez-Rey E, Rico L Treatment of experimental arthritis by inducing immune tolerance with human adipose-derived mesenchymal stem cells. Arthritis Rheum . 2009; 60: 1006– 1019 Google Scholar CrossRef Search ADS PubMed  41. Planat-Benard V, Silvestre JS, Cousin B Plasticity of human adipose lineage cells toward endothelial cells. Circulation . 2004; 109: 656– 663 Google Scholar CrossRef Search ADS PubMed  42. Ebrahimian TG, Pouzoulet F, Squiban C Cell therapy based on adipose tissue-derived stromal cells promotes physiological and pathological wound healing. Arterioscler Thromb Vasc Biol . 2009; 29: 503– 510 Google Scholar CrossRef Search ADS PubMed  43. Mazo M, Planat-Bénard V, Abizanda G Transplantation of adipose derived stromal cells is associated with functional improvement in a rat model of chronic myocardial infarction. Eur J Heart Fail . 2008; 10: 454– 462 Google Scholar CrossRef Search ADS PubMed  44. Andersen DC, Schroder HD, Jensen CH Non-cultured adipose-derived CD45- side population cells are enriched for progenitors that give rise to myofibres in vivo. Exp Cell Res . 2008; 314: 2951– 2964 Google Scholar CrossRef Search ADS PubMed  45. Meier JKH, Scharl M, Miller SN Specific differences in migratory function of myofibroblasts isolated from Crohn's disease fistulae and strictures. Inflamm Bowel Dis . 2011; 17: 202– 212 Google Scholar CrossRef Search ADS PubMed  Copyright © 2017 Crohn's & Colitis Foundation
TI - Comparison of Stromal Vascular Fraction with or Without a Novel Bioscaffold to Fibrin Glue in a Porcine Model of Mechanically Induced Anorectal Fistula
JF - Inflammatory Bowel Diseases
DO - 10.1097/MIB.0000000000001254
DA - 2017-11-01
UR - https://www.deepdyve.com/lp/oxford-university-press/comparison-of-stromal-vascular-fraction-with-or-without-a-novel-0DheOUJ0CO
SP - 1962
EP - 1971
VL - 23
IS - 11
DP - DeepDyve
ER -