W ith unrelieved pressure, tissue ischemia develops, and metabolic wastes accumulate in the interstitial tissue, resulting in anoxia and cellular death.1 This pressure-induced ischemia also leads to excessive tissue hypoxia, further promoting bacterial proliferation and tissue destruction.2 In clinical settings, some pressure-induced ischemic wounds are caused by support surfaces, and some lower-limb diabetic pressure ulcers are caused by footwear. Infection of these wounds considerably impairs the healing process. For example, with pressure ulcers, the prevalence of infection is 1.4 per 1,000 ulcer days3 or 27% of studied pressure ulcers.4 Moreover, the mortality rate for bacteremia due to pressure ulcer infection has been reported at 15.4%.5 In another study, 48% of patients with pressure ulcers died of sepsis.6 The results of the aforementioned studies suggest that if a bacterial infection develops in a pressure-induced ischemic wound, medical conditions can easily deteriorate and possibly become life threatening. Although the relationship between bacteria and acute wounds using animal models has been closely examined at the postoperative stage,7–13 no such animal model has been studied to examine the relationship between bacteria and pressure-induced ischemic wounds.1 The purpose of this study was to develop an animal model of pressure-induced ischemic wounds and to clarify the effects of bacteria on the healing process of such wounds. Methods Subjects. Fifty-nine male Wistar rats were used. Unlike humans, rats are loose-skinned animals. The properties of loose skin allow wound contraction to play a significant role in closing rat skin wounds.14 While the rat may not be the ideal model of pressure-induced ischemic wounds in humans, the authors decided to use this species, since it has been used previously for many wound models.14 Protocol. Wound healing was compared macroscopically and histologically between a group of rats with pressure-induced ischemic wounds inoculated with bacteria (bacterial-inoculation group) and a group of rats with pressure-induced ischemic wounds without bacterial inoculation (control group). To ensure reproducibility, wounds were experimentally created on the flank regions where vertical pressure can be applied for a set period of time with a degree of stability. Instrumentation. The authors’ original experimental device was constructed from a pressure applicator and an air compressor to artificially create pressure-induced ischemic wounds (Figure 1). Compressive pressure was applied by lowering the cylinder. A pressure sensor (strain gauge) was placed between the cylinder and the pressure applicator, and the results were displayed using a digital indicator (CSD-701, Minebea Co. Ltd., Japan). The position of the cylinder could be adjusted manually to vary the amount of pressure. The end of the pressure applicator was circular in shape with an area of 3 cm2. A small air compressor (PA400B, Hitachi, Japan) was used to lower the pressure applicator. Preparation of ischemic wounds. Using the experimental device, 8 kg of pressure was applied for 6 hours. Because previous studies used different animals and experimental devices, the authors conducted a pilot study involving 24 rats to clarify the necessary amount and duration of pressure to prepare pressure-induced ischemic wounds using an experimental device. The intensity and duration of pressure used here were decided based on a pressure-time curve obtained from the authors’ pilot study. Bacterial inoculation. Bacterium. A clinical isolate of Staphylococcus aureus was used, because this bacterium is often detected in human chronic wounds.15 The bacterial strain used in the present study was subcultured every week using special agar media. A single colony was collected using a sterile platinum loop, placed in 3 mL of sterile 2.5% heart infusion broth (Nissui, Japan), and cultured for 18 hours in a 37oC incubator. Aliquots of 0.1 mL of the resulting solution with viable cell counts of 5.3 x 107/mL were used for inoculation. Inoculation routes. In another preliminary study involving 9 rats, the most effective route of bacterial inoculation was determined. Staphylococcus aureus was inoculated into the ischemic wound model through 4 routes: topical application to the epidermis (n = 2), subcutaneous injection (n = 2), intramuscular injection into the abdominal muscle (n = 2), and all 3 layers (“all layers;” epidermis, subcutaneous, and abdominal muscle, n = 2). In a control rat, the same level of inoculum was applied in the same way as for the all layers group without prior application of pressure. The results were evaluated macroscopically using the DESIGN Tool16 (Table 1). Wound healing was found to be slowest when the bacterium was inoculated into the epidermis, subcutaneous tissue, and abdominal muscle. Figure 2 shows images of the bacteria observed in the muscle layer and subcutaneous tissue. By Day 14 in the all layers rats, the lesion extended into the dermis of 1 rat, and its depth could not be measured. For the other rat, the lesions extended to the intramuscular tissue. The control inoculation site appeared normal from Days 0 to 14 with no color change or exudation exhibited. On Day 14, histological analysis showed no epidermal loss, and the structure and alignment of dermal collagen fibers and the distribution of integumentary appendages were normal. No degeneration of the dermal muscle, subcutaneous tissue, or muscular tissue was seen. In addition, while a diffuse infiltrate of lymphocytes was observed in the connective tissue, tissue degeneration was not apparent. Staphylococcus aureus was not detected. Procedures. After weighing and etherizing each rat, 30 mg/kg body weight of sodium pentobarbital (Nembutal®, Abbott, Japan) was administered intraperitoneally, and the right flank region was shaved using an electric clipper. Two 2-cm incisions, 5 cm apart, that extended to the peritoneum were then made in the right flank region using electrocautery. A 2-cm-wide metal plate was inserted from 1 incision and exited from the other. The rat and metal plate were then fixed to the experimental device. Another preliminary study had been conducted to ascertain the effects of flank incisions on the blood circulation. Under general anesthesia, after performing incisions as described previously, the chest was opened, and a catheter was inserted from the left atrium to the aorta to inject dye. The spread of the dye to the normal left abdominal sides and the right abdominal sides with 2 incisions was examined; the results showed that both sides were stained in the same manner, thus suggesting that the incisions did not affect the blood circulation in the flank skin. In the bacterial-inoculation group, each rat was fixed to the experimental device, and 0.1 mL of the cultured Staphylococcus aureus (5.3 x 107 CFU/mL) in 2.5% heart infusion broth solution was inoculated in the following manner: after placing 4 drops on the epidermis, the remaining solution was injected intramuscularly and subcutaneously. Inoculation was performed using a 1-mL disposable syringe with a 27-gauge, 19-mm length needle. Solution was injected subcutaneously at a 45-degree angle with half the length of a needle and intramuscularly at a 90-degree angle with two-thirds length of a needle at the center of the compressed area. The inoculated area was then covered with a sterile transparent sheet of hydrocolloid dressing (Tegasorb, 3M Health Care, St Paul, Minn) (3 cm x 3 cm), and 8 kg of pressure were applied for 6 hours. After applying the compressive pressure, bacterial inoculation was repeated in the same manner. The metal plate was removed, and the incisions were sutured. Thirty minutes after the completion of pressure application, the compressed area was observed and photographed. The entire right flank region was then covered using a 5 cm x 10 cm hydrocolloid dressing (Tegasorb, 3M Health Care) to prevent the animals from licking the inoculated area of the epidermis and to protect the incision wounds. The rats were then placed in separate cages. The skin on the compressed area was observed and photographed every day. The incision wounds were cleaned daily using physiological saline. In the control group, the same procedures were conducted with the exception of bacterial inoculation. Analysis. Skin changes were assessed in terms of color and wound depth at 30 minutes and at 1, 3, 7, and 14 days after the completion of pressure application. Macroscopic examinations were performed using 6 rats (3 from each group). The DESIGN classification system16 was used to assess invasion depth. In addition, epidermal loss, edema, cellular damage, vascular changes, bleeding, collagen fiber alignment and thickness, and inflammatory cell infiltration were qualitatively analyzed under light microscopy. Altered collagen fiber alignment and thickness were assessed by comparing the areas of pressure application to the epidermis where no pressure was applied. Histological examinations were conducted with 20 rats using 2 rats at each time interval per group. Histological findings were evaluated using qualitative variables in consensus method, and no statistical analysis was conducted. Tissue samples used in histological examinations were fixed in 10% formaldehyde buffer solution, dehydrated using 95% and then 100% ethanol, cleaned in xylene, and embedded in paraffin. Next, 5mm sections were prepared, deparaffinized, and stained using hematoxylin-eosin (HE). Histological examinations were not conducted in a blind manner because of the presence of inflammatory reactions unique to bacterial invasion. Ethical considerations. The study protocol was approved by the Kanazawa University Takaramachi District Animal Study Committee. When a rat awoke from anesthesia during pressure application, pentobarbital sodium was administered intraperitoneally to alleviate pain. Rats were placed in separate cages at 24 ± 1oC with a relative humidity of 55 ± 5%. During the study, animals had free access to tap water and pelleted food. Histopathological samples were collected by administering a lethal dose of pentobarbital sodium intraperitoneally and surgically removing the compressed area with a 1-cm margin. Rats were euthanized once the tissue samples were harvested. Results Healing process of the ischemic wound (control group). Macroscopic findings (Figure 3). At 30 minutes after pressure application, a dark red circle matching the compressed area with edema at the circumference was seen. By Day 1, this circle had turned white, and on Day 3, it had turned pink, by which time the volume of exudate had significantly increased. By Day 7, the pink circle had turned white and had developed several bright red spots. By Day 14, the entire compressed area had become glossy white. Histological findings. Up to Day 3, the epidermis was absent, but by Day 7, the wound was covered by thickened epidermis. In the dermal layer, on Day 1, follicular necrosis, vasodilatation, and red blood cell extravasation were observed. On Day 3, follicular necrosis and red blood cell extravasation were present, and the degree of vasodilatation had become more severe. On Day 7, follicular necrosis was seen in some areas. On Day 14, histological findings similar to those of normal tissue were observed (Figure 4). Throughout the observation period, the alignment and structure of collagen fibers were normal (Table 2a and Figure 4). In the dermal muscle, on Day 1, muscle fiber degeneration and inflammatory cell infiltration among muscle fibers were confirmed. On Day 3, muscular degeneration, necrosis, and inflammatory cell infiltration were the most severe, and red blood cell extravasation was observed. On Day 7, muscular degeneration and necrosis were less severe than on Day 3, but inflammatory cell infiltration remained. On Day 14, the degree of inflammatory cell infiltration had improved (Table 3a). Subcutaneous tissue is composed mostly of fat. In this layer, extensive necrosis and vasodilatation were seen up to Day 3. Inflammatory cell infiltration was observed starting on Day 1 and was most notable on Day 7, when numerous fibroblasts were observed and granulation tissue formation was seen (Table 4a). In the abdominal muscle layer, histological damage was marked on Day 3 and began to improve after Day 7 (Table 5a). Healing process of ischemic wounds with bacterial inoculation (bacterial-inoculation group). Macroscopic findings. At 30 minutes after the experiment, a dark red circle matching the pressure application area had formed. This had turned white by Day 1. On Day 3, the pressure application area was brown at the center and exhibited a color gradation from yellow to pink, and the edge was red. The volume of exudate had also increased significantly by this stage. On Day 7, thick yellowish-white necrotic tissue had formed over the entire pressure application area, and on Day 14, the entire necrotic tissue was on the verge of being sloughed from the wound (Figure 5). Histological findings. The epidermis did not repair throughout the observation period. On Day 14, the wound surface was covered by necrotic tissue. In the dermal layer, on Day 1, follicular necrosis, vasodilatation, red blood cell extravasation, and zonal neutrophilic infiltration were observed. On Day 3, collagen fibers were disrupted. On Day 7, collagen fibers were necrotic, and inflammatory cell infiltration was seen throughout most of the dermal layer (Figure 6). By Day 14, the dermal layer had been obliterated (Table 2b). In the dermal muscle, on Day 1, zonal inflammatory cell infiltration and muscular degeneration were confirmed, and on Days 3 and 7, muscular degeneration and necrosis became more remarkable. By Day 14, the dermal muscle layer had been obliterated (Table 3b and Figure 7). In the fat layer (subcutaneous tissue), on Day 1, zonal neutrophilic infiltration and marked cellular degeneration and necrosis were seen. On Day 3, many fibroblasts were observed, and on Day 7, granulation tissue formation with prominent neoangiogenesis was seen in the wound margin. By Day 14, the wound exhibited no fat tissue (Table 4b). In the abdominal muscle layer, on Day 1, zonal inflammatory cell infiltration, muscular degeneration, and necrosis were observed. On Day 3, muscular degeneration and necrosis were seen, but the continuity between the abdominal muscle layer in the wound and the healthy abdominal muscle layer on both ends was maintained. However, on Day 7, the healthy abdominal muscle layer was physically separated from the muscle layer in the wound by granulation tissue, and the abdominal muscle layer of the wound was isolated without being necrotic. On Day 14, the isolated abdominal muscle layer in the wound was surrounded by U-shaped granulation tissue and appeared to be being pushed upward (Table 5b and Figure 7). Bands consisting of numerous neutrophils and macrophages were the specific histological findings in the bacterial-inoculation group. On Day 1, 2 bands—1 reaching from the epidermis to the fat layer and another from the fat layer to the abdominal muscle—sandwiched the pressure application area. By Day 3, these 2 bands had expanded from the epidermis to the deep muscle layer. On Day 7, a U-shaped band was seen. The upper layer of the inner section of this band consisted of necrotic tissue, while the lower section consisted of granulation tissue. On Day 14, a thick abscess membrane rich in vessels was seen at the outer margin of the U-shaped inflammatory cell zone (Figure 6). Discussion To the authors’ knowledge, this is the first observation suggesting that the subcutaneous presence of a pathogenic strain of bacteria increases tissue susceptibility to chronic ulceration in response to pressure-induced ischemia. This finding raises several interesting questions for future research. For example, what other strains or species of bacteria or other organisms potentiate ulceration in this way? How would local or systemic antibiotics or other antimicrobial agents affect this interaction? How would skin maceration alter the likelihood or consistency of ulceration? What are the clinical implications for post-surgical care of bony prominences subjected to prolonged pressure in the operating room and subsequent inoculation via body fluids? Human wound healing is a process that is significantly affected by systemic conditions. In the present study, the authors focused on pressure-induced ischemia and local infection, which are major factors contributing to the delayed healing of wounds. Although animal models of ischemic wounds with surgical restriction of blood exist, an animal model involving pressure-induced ischemic wounds has yet to be developed. Due to the chronic action of pressure, ischemia of wound tissue occurs primarily in patients with vascular disease, diabetes, and immobilization. Although no 2 ischemic wounds are the same, an animal model of pressure-induced ischemic wounds would be useful in advancing knowledge of factors affecting deterioration and healing of such clinical wounds. Furthermore, such an animal model would allow exploration of the effects of these contributing factors separately and as they interact to impair wound healing. Validity of the pressure-induced ischemic wound model. When experimentally inducing an ischemic wound, blood circulation can be obstructed by vessel ligation17 or skin banding.18 In both of these techniques, a wound is prepared by directly blocking the skin circulation, and as a result, the circulation in deep tissue remains normal. However, in the pressure-induced ischemic wounds, the direct pressure application to the epidermis subjects the deep tissue vessels to occlusion and reduction in blood flow, which results in deep tissue damage. In this regard, the methods in the present study, in which pressure is applied vertically to the skin, more closely represent pressure-induced ischemic wounds. In a previous study18 of a rat model of pressure ulcers, greater damage occurred in the deep tissue than in the epidermis. In the present control group, those histological findings were also observed; thus, the validity of the authors’ model is supported histologically. The authors inserted a metal plate to apply a certain amount of pressure for a set length of time. It has been noted that when pressure is applied to the dermal tissue on the bone, degeneration of the deep muscle layer occurs,19,20 which accords with the findings for the muscle layer in the control group in the present study. The authors, therefore, considered the effects of the insertion of the metal plate to be minimal. Moreover, the authors’ preliminary dye injection experiment confirmed that the incisions did not affect the blood circulation in the flank region. Relationship between bacterial inoculation and healing of pressure-induced ischemic wounds. To consider the relationship between bacterial inoculation and healing of pressure-induced ischemic wounds, the authors searched the literature for studies in which bacteria were inoculated into full-thickness wounds. Tachi and colleagues13 prepared full-thickness wounds in rats using a puncher and then inoculated Staphylococcus aureus into the wounds (1.0 x 107 CFU/wound). They reported that while serous or purulent exudate persisted, granulation tissue formation was seen 3 days after wound development. These findings show that in rats, even when bacteria are inoculated into full-thickness wounds, abscess formation does not occur unless the wounds are ischemic. Concerning the healing process of pressure-induced ischemic wounds without bacterial inoculation (control group), the epidermis, subcutaneous tissue, and abdominal muscle were damaged, but granulation tissue formation was seen on Day 7, and the wound had healed without ulceration by Day 14. Due to the high tolerance of the collagen fibers for the ischemic conditions created in this study, full-thickness wounds did not occur. Conversely, in the healing process of pressure-induced ischemic wounds with bacterial inoculation (bacterial-inoculation group), bacterial proliferation was noted, and immunological reactions against the infection resulted in abscess formation in several sites from the epidermis to the muscle layer. Thus, full-thickness wounds did occur, and wound healing was delayed. In general, when bacteria invade tissue, the complement system is activated, and potent chemotactic factors for phagocytic cells are produced to enable microbes to be engulfed. The bacterial-inoculation group resembled the control group from Days 1–3 in terms of neutrophil infiltration and collagen fibers within the dermal layer. However, on Day 7, thick necrotic tissue matching the pressure application area was observed macroscopically, and dermal layer necrosis was seen histologically. Hence, after Day 3, the healing process differed between the bacterial-inoculation group and the control group. In other words, in the bacterial-inoculation group, necrosis of the dermal layer (which provided high tolerance to ischemia) began after Day 3. The pressure application persistently disrupted the blood circulation in the muscle layer and subcutaneous tissue, which resulted in hypoxia. This persistent hypoxia is likely to facilitate bacterial proliferation, suppress neutrophil apoptosis, and release proteases to induce tissue injury, ie, necrosis of the dermal layer. Additionally, the authors noted unique findings associated with ischemic wounds inoculated with bacteria. Granulation tissue (abscess membrane) had separated the abdominal muscle layer in the wound from the marginal healthy abdominal muscle by Day 7, and the ischemic wound area was elevated on Day 14. In the control group, degeneration and necrosis of the abdominal muscle layer were observed, but muscle layer isolation in the wound and tissue bulging were absent. This fragmentation appears to represent an immunological reaction to prevent the bacterial proliferation, while the bulging is considered a result of marked wound contraction in attempt to eliminate the abscess. As to the relationship between bacteria and wound contraction, it is reported that contraction in acute wounds is delayed in the presence of bacterial infection;21 however, the relationship between bacterial proliferation and contraction of pressure-induced ischemic wounds has not been clarified. The mechanism underlying the contraction of pressure-induced ischemic wounds inoculated with bacteria needs to be ascertained in future studies. A limitation of the present study is that only 1 type of bacterium was used; therefore, it is not clear if the same results can be obtained using other bacterial species. Further, although the inoculation was quantified in terms of viable cell counts, the authors did not measure the amount of bacterial proliferation in wounds during abscess formation. Future studies need to ascertain the level of bacterial proliferation in ischemic wounds. Despite these limitations and the small number of wounds observed at each time point, the consistency of these observations in this animal model over time lends strength to the hypothesis that bacterial invasion may interact with pressure-induced ischemia to generate ulceration. Conclusions Ischemia alone did not suffice to create enduring full-thickness damage to rat skin subjected to prolonged pressure of 8 kg for 6 hours. Inoculation at any 1 tissue level with 5.3 x 107/mL of Staphylococcus aureus into the pressure-induced ischemic zone also failed to induce consistent ulceration. Only when this inoculum was applied to the epidermis, subcutaneum, and underlying muscle before and after pressure-induced ischemia were abscess formation, ulceration, and necrosis evident at 14 days. As a result, healing of the ischemic wound may be delayed, because the dermal layer collagen fibers that resist ischemia became necrotic due to bacterial proliferation and neutrophilic infiltration.