Role of Growth Factors in the Treatment of Diabetic Foot Ulceration

20 Role of Growth Factors in the Treatment of Diabetic Foot Ulceration David L. Steed, MD PRINCIPLES OF WOUND HEALING AND GROWTH FACTOR THERAPY Wound healing is the process of tissue repair and the tissue response to injury. It is a complex biological process involving chemotaxis, cellular reproduction, matrix protein production and deposition, neovascularization, and scar remodeling (1). Growth factors are polypeptides that control the growth, differentiation, and metabolism of cells, and regulate the process of tissue repair (2,3). The role of growth factors in wound healing and specifically in diabetic ulcer healing is the subject of this chapter. The three phases of wound healing inflammation, fibroplasia, and maturation are each controlled by growth factors that are present in only small amounts, yet powerful in their influence on wound repair. There is great interest in manipulating the cellular environment of the wound with proteins, growth factors, and gene therapy. The first phase of wound healing is the inflammatory response, initiated immediately after the injury (4). Vasoconstriction limits hemorrhage to the site of wounding. As blood vessels are damaged and blood leaks from within the lumen, platelets come into contact with collagen in the wall of the vessel beneath the endothelium. Platelets are activated by the collagen and initiate coagulation. Serotonin and thromboxane are released and enhance vasoconstriction locally, keeping the healing factors within the wound. Simultaneously, vasodilatation occurs, allowing new factors to be brought into the wound. Vasodilatation is mediated by histamine, released by the platelets, mast cells, and basophils. There is also an increase in vascular permeability allowing blood borne factors to enter this area. Arachidonic acid is produced and serves as an intermediate for production of prostaglandins and leukotrienes. These proteins are intense vasodilators which increase vascular permeability, along with histamine, bradykinin, and complement. Thromboxane also increases platelet aggregation and local vasoconstriction. Platelets control hemorrhage by initiating clotting through the coagulation system. The intrinsic system is activated by Hageman factor, known as factor XII, as it comes into contact with collagen. In the presence of kininogen, a precursor of bradykinin and prekallikrein, factor XII activates factor XI, then factor IX, and then factor VIII. Thromboplastin triggers a response in the extrinsic system. Thromboplastin is formed as phospholipids and glycoproteins are released by blood coming into contact with the injured tissues. Factor VII is activated in the presence of calcium. Both the intrinsic and From: The Diabetic Foot, Second Edition Edited by: A. Veves, J. M. Giurini, and F. W. LoGerfo Humana Press Inc., Totowa, NJ 447

448 Steed extrinsic systems activate the final common pathway, producing fibrin and leading to fibrin polymerization. To balance the coagulation cascade, the fibrinolytic system is activated. This system monitors clotting to prevent coagulation from extending beyond the wound. It is activated by the same factors that initiate coagulation and thus regulates the process. The complement cascade is activated by platelets and neutral proteases. This system produces potent proteins known as anaphylotoxins, which cause mast cells to degranulate and release histamine. Substances released by the inflammatory process are chemoattractants for neutrophils. These cause margination of white blood cells and then migration of these white blood cells into the wound. The neutrophils are phagocytes for bacteria. Wounds can heal without white blood cells, but the risk of infection is increased. Neutrophils produce free oxygen radicals and lysosomal enzymes for host defense. The neutrophils are later removed from the wound by tissue macrophages. Monocytes enter the wound space and become tissue macrophages. They take over control of the wound environment by the third day. Wounds cannot heal without the macrophage. These cells regulate the production of growth factors including plateletderived growth factor (PDGF), tumor necrosis factor (TNF), and transforming growth factor (TGF)-β; thus they control protein production, matrix formation, and remodeling. Extracellular matrix (ECM) is a group of proteins in a polysaccharide gel made up of glycosaminoglycans and proteoglycans produced by the fibroblast. These proteins are structural such as collagen and elastin, or are involved in controlling cell adhesion such as fibronectin and laminin (5). Thrombospondin and von Willebrand factor are other adhesion molecules. Fibronectin is also a chemoattractant for circulating monocytes and stimulates its differentiation into tissue macrophages. The second phase of wound healing is fibroplasia and begins with macrophages and fibroblasts increasing in number in the wound, whereas white blood cells decrease as fewer enter the wound. The inflammatory response ends as the mediators of inflammation are no longer produced, and those already present are inactivated or removed by diffusion or by macrophages. Fibroplasia begins around the fifth day following injury and may continue for 2 weeks. This begins the process of matrix formation, especially collagen synthesis. Angiogenesis is the process of rebuilding the blood supply to the wound (6). Fibroblasts are attracted to the wound and replicate in response to fibronectin, PDGF, fibroblast growth factor (FGF), TGF-β, and C5a, a product of the complement system. Fibroblasts produce proteoglycans and structural proteins. The cellular matrix is made up of hyaluronate and fibronectin which allow for cellular migration through chemotactic factors formed in the wound. Fibronectin binds proteins and fibroblasts in the matrix and provides a pathway along which fibroblasts can move. Fibronectin also plays a role in epithelialization and angiogenesis. Collagen is the most common protein in the mammalian world and is produced by the fibroblast. It is a family of at least 12 proteins, rich in glycine and proline, and bound in a tight triple helix. Cross-linking between the three strands of collagen provides for a very stabile molecule, resistant to breakdown. Macrophages control the release of collagen from fibroblasts through growth factors such as PDGF, epidermal growth factor (EGF), FGF, and TGF-β. Collagen is remodeled for several years in a healing wound. Elastin is the other major structural protein and contains proline and lysine. It is present as random

Role of Growth Factors in Diabetic Foot Ulceration 449 coils, allowing both stretch and recoil. It is present in much smaller amounts than collagen. Angiogenesis occurs by the budding of existing capillaries after stimulation by FGF. Endothelial cells proliferate and migrate through the healing wound, allowing connections between the capillaries to form a vascular network in the wound space. This capillary network provides an avenue of access for new healing factors into the wound and ends when the wound has an adequate blood supply. Hypoxia triggers angiogenesis; thus it appears as if this process if controlled by oxygen tension (7,8). Epithelialization occurs as cells migrate from the edge of the wound over a collagenfibronectin surface. This process results in mature skin covering the wound. Scar contracture then occurs as the wound matures. The final phase of wound repair is maturation or scar remodeling. Wound remodeling involves a number of proteins including hyaluronidase, collagenase, and elastase. Hyaluronate in the matrix is replaced by dermatan sulfate and chondroitin sulfate. These proteins reduce cell migration and allow cell differentiation. Plasmin, which is formed from plasminogen, degrades fibrin. Urokinase, produced by leukocytes, fibroblasts, endothelial cells, and keratinocytes, activates collagenase and elastase. Collagenase, which allows collagen remodeling, is secreted by macrophages, fibroblasts, epithelial cells, and white blood cells. It is able to break the collagen triple helix to allow remodeling. The scar becomes less hyperemic and less red in appearance as blood supply is reduced. The scar remodels and wound strength increases for up to 2 years following injury, yet the total collagen content of the wound does not change. GROWTH FACTORS Growth factors are polypeptides that initiate the growth and proliferation of cells and stimulate protein production (2,3). They are named for their tissue of origin, their biological action, or the cell on which they exert their influence. Growth factors may have paracrine or autocrine function whereby they affect not only adjacent cells but also have a self-regulating effect. Some are transported plasma bound to large carrier proteins and thus serve an endocrine function. They are produced by a variety of cells including platelets, macrophages, epithelial cells, fibroblasts, and endothelial cells. Growth factors are chemoattractants for neutrophils, macrophages, fibroblasts, and endothelial cells. Growth factors bind to specific receptors on the cell surface to stimulate cell growth. Although present in only minute amounts, they exert a powerful influence on wound repair. The growth factors involved in wound healing include PDGF, TGF-β, EGF, FGF, and insulin-like growth factor (IGF; Table 1). The platelet that is critical to the initiation of the wound-healing process is rich in growth factors. Growth factors initially released in the wound space by platelets are subsequently degraded by proteases. Other cells that have been drawn into the wound space such as inflammatory cells, fibroblasts, and epithelial cells are also involved in growth factor production. Macrophages also release factors such as TNF. Keratinocytes are stimulated by EGF, IGF-1, TGF-α, and interleukin (IL)-1. Wound remodeling occurs under the control of collagenase, produced in response to EGF, TNF, IL-1, and PDGF. Thus, all of wound healing is under the direct or indirect control of growth factors. It is reasonable then to speculate that exogenous growth factors applied to the wound may influence healing.

450 Steed Table 1 Growth Factors, Their Sources and Actions Growth factor Cell of origin Action PDGF Platelets Chemoattractant and mitogen for fibroblasts, smooth muscle cells, macrophages, inflammatory cells. TGF-α Endothelium Fibroblasts Macrophages Mitogens for keratinocytes and fibroblasts. Keratinocytes Angiogenesis factor. Hepatocytes Eosinophiles TGF-β Platelets Chemoattractant for neutrophils, macrophages, fibroblasts. Stimulates collagen synthesis, granulation tissue formation, reepithelialization. Macrophage Fibroblasts Keratinocytes Lymphocytes VEGF Endothelial cells Potent mitogen for endothelial cells EGF Platelet Promotes epidermal regeneration. Increases fibroblasts in a wound. Kidney Salivary gland Lacrimal gland FGF Fibroblasts Mitogen for endothelial cells. Angiogenesis factor. Endothelial cells Smooth muscle cells Chrondrocytes KGF Fibroblasts Stimulates keratinocytes. IGF Liver Stimulate synthesis of glycogen and transport of glucose and amino acids across cell membranes. Increase collagen synthesis by the fibroblast. Lung Heart Lung Pancreas Brain Muscle PLATELET-DERIVED GROWTH FACTOR PDGF has been studied more widely than any other growth factor and is approved for clinical use. PDGF has a molecular weight of 24,000 Daltons. It is a potent chemoattractant and mitogen for fibroblasts, smooth muscle cells, and inflammatory cells. PDGF is produced by platelets, macrophages, vascular endothelium, and fibroblasts (9). It is made up of two chains, an A and a B chain, held together by disulfide bonds in three

Role of Growth Factors in Diabetic Foot Ulceration 451 dimeric forms, AA, AB, and BB. There is a 60% amino acid homology between the two chains. Human platelets contain all three forms of PDGF in a ratio of about 12% AA, 65% AB, and 23% BB. The B chain is quite similar to the transforming gene of the simian sarcoma virus, an acute transforming retrovirus. The human proto-oncogene C-sis is similar the viral oncogene V-cis and encodes for the B chain of PDGF. There are two PDGF receptors, α- and β-receptor. The α receptor recognizes both the A and B chains of PDGF and thus can bind to the AA form, the AB form, and the BB form. The β receptor recognizes only the B chain, and thus binds only to the BB form and weakly to the AB form. Most cells have many times more β-receptors than α-receptors. Cells with PDGF receptors include fibroblasts, vascular smooth muscle cells, and some microvascular endothelial cells. PDGF acts with TGF-β and EGF to stimulate mesenchymal cells. Although PDGF is produced by endothelial cells of the vascular system, the endothelial cells do not respond to PDGF. Rather, they work in a paracrine manner to stimulate adjacent smooth muscle cells. Smooth muscle cells also act in an autocrine fashion and produce PDGF. PDGF is stabile to extremes of heat, a wide range of ph, and degradation by proteases. The principle cells involved in the early stages of wound healing all synthesize and secrete PDGF. Platelets, among the first cells to enter the wound, are the largest source of PDGF in the human body. Circulating monocytes are attracted to the wound and become tissue macrophages. These cells also produce PDGF. PDGF stimulates the production of fibronectin and hyaluronic acid, proteins which are important components of provisional matrix. Collagenase, a protein important in wound remodeling, is also produced in response to PDGF. There are no reported cases of a human deficient in PDGF suggesting that PDGF is critical to the survival of the individual. PDGF has been manufactured by recombinant DNA technology. In animal models, it has been shown to improve the breaking strength of incisional wounds when applied topically as a single dose. It also accelerated acute wound healing. By 3 months, however, there was no difference in wound healing as compared with untreated wounds suggesting that although PDGF accelerated wound healing, the wound healing was quite similar to normal healing. Wounds treated with PDGF had a marked increase in inflammatory cells entering the wound, including neutrophils, monocytes, and fibroblasts. As a result of this cellular response, granulation tissue production was also increased. Despite the fact that PDGF does not directly affect keratinocytes, wounds in animals were shown to have an increased rate of epithelialization. This is probably owing to the influence from macrophages and fibroblasts attracted into the wound space by PDGF. Wounds treated topically with PDGF have an increase in neovascularization, although PDGF does not directly stimulate endothelial cells. Thus, it appears as if PDGF accelerates wound healing by accelerating the normal sequences of healing. The healed wounds appear to be normal in all aspects. PDGF has been studied extensively in clinical trials. The effectiveness of recombinant human PDGF-BB in healing was first studied in decubitus ulcers (10,11). Patients were treated with PDGF topically and followed for 28 days. There was a greater amount of wound closure in patients treated with the highest dose of PDGF. The lower doses had little effect. In another trial, patients with decubitus ulcers were treated with 100 or 300 µg/ml or placebo again for 1 month. The ulcer volume was significantly

452 Steed reduced in the PDGF-treated patients. No significant toxicity related to PDGF was noted. Complete wound closure was not an end point in either study and thus the question regarding whether PDGF could accomplish complete wound healing in humans was not answered. A randomized prospective double blind trial of recombinant PDGF-BB was performed in patients with diabetic neurotrophic foot ulcers (12). Patients were treated with PDGF at a dose of 2.2 µg/cm 2 of wound in vehicle, carboxymethylcellulose, or vehicle alone for 20 weeks or until complete wound closure occurred. Patients had wounds of at least 8-week duration, were considered to be free of infection, and had an adequate blood supply as demonstrated by a transcutaneous oxygen tension (TcPO 2 )of at least 30 mmhg. All wounds were debrided by completely excising the wound prior to entry into the study and as needed during the trial. In these patients with chronic nonhealing wounds, 48% healed following treatment with PDGF, whereas only 25% healed with vehicle alone (p < 0.01). The median reduction in wound area was 98.8% for PDGF-treated patients but only 82.1% for those treated with vehicle. There were no significant differences in the incidence or severity of adverse events in either group. This was the first clinical trial to suggest that a growth factor, PDGF, could be applied topically and be effective and safe in accelerating the healing of chronic wounds in humans. In another trial using recombinant human PDGF in the treatment of similar patients with diabetic foot ulcers, those patients treated with PDGF-BB had an increase in the incidence of complete wound closure of 43% as compared with placebo (p = 0.007). PDGF also decreased the time to achieve complete wound closure by 32% (p = 0.013) when compared with the placebo. In reviewing patients treated with PDGF or vehicle alone, it was noted that those patients receiving the best wound care healed better whether PDGF or vehicle was applied. Debridement proved to be critically important. The benefits from PDGF will be minimized if the wounds are not treated properly. The vehicle, carboxymethylcellulose, was tested to determine if it was inert in wound healing. It did provide a moist environment for wound healing but did not improve healing significantly. PDGF is approved for use in the United States and is sold as Regranex. TRANSFORMING GROWTH FACTOR The growth factor studied most extensively after PDGF is TGF-β. Transforming growth factors are made up of two polypeptide chains, α and β. TGF-α has a 30% amino acid homology with EGF. It is named because of its ability to reversibly stimulate the growth of cells. Cancer cells do this also. TGF-α is produced by many different cells including macrophages, keratinocytes, hepatocytes, and eosinophiles (13). TGF-α and EGF are mitogens for keratinocytes and fibroblasts, but TGF-α is a more potent angiogenesis factor. Both TGF-α and EGF bind to the EGF receptor but their specific actions may be different partly owing to differences in their binding. As yet there have been no clinical trials of wound healing with TGF-α. TGF-β has no amino acid homology with TGF-α or any other group of growth factors. TGF-β is a group of proteins which can reversibly inhibit growth of cells especially those of ectodermal origin. TGF-β is structurally similar to bone morphogenic protein.

Role of Growth Factors in Diabetic Foot Ulceration 453 TGF-β is produced by a variety of cells including platelets, macrophages, fibroblasts, keratinocytes, and lymphocytes. TGF-β is a chemoattractant for neutrophils, macrophages, and fibroblasts and stimulates collagen synthesis, granulation tissue formation, and reepithelialization (14,15). Nearly all cells have receptors for TGF-β and have the potential to respond to it. TGF-β can stimulate or inhibit the growth or differentiation of many different cells. It appears as if TGF-β is the most widely acting group of growth factors. There have been three forms of TGF-β isolated, TGF-β 1, TGF-β 2, and TGF-β 3. The actions of the three different forms of TGF-β are very similar. TGF-βs have a molecular weight of about 25,000. They reversibly stimulate growth of fibroblasts and thus received their name. TGF-βs are potent stimulators of chemotaxis in inflammatory cells and trigger cells to produce ECM. Thus, the TGF-β is important in wound healing. TGF-β may be associated with scarring. Increased TFG-β has been found in fibroblasts of human hypertrophic burn scar and in keloids (16 19). TGF-β was tested in diabetic wounds in three different doses in a collagen sponge (20). There was also a standard care arm in the trial. There was a significant improvement in the healing of diabetic ulcers treated with each of the three doses of TGF-β compared with collagen sponge. In this study, the patients treated with standard care healed better than those patients treated with TGF-β. The role of TGF-β in treating patients with chronic wounds remains undetermined. VASCULAR ENDOTHELIAL GROWTH FACTOR Vascular endothelial growth factor (VEGF) is quite similar to PDGF. It has a molecular weight of 45,000 Daltons. It has a 24% amino acid homology to the B chain of PDGF. VEGF, however, binds different receptors than PDGF and has different actions. Although it is a potent mitogen for endothelial cells, it is not a mitogen for fibroblasts or vascular smooth muscle cells as is PDGF. VEGF is angiogenic and may play a role in wound healing by way of this property. VEGF antibodies reduce angiogenesis and granulation tissue formation (21). There has been interest in using VEGF to stimulate the development of collateral arteries in patients with vascular disease, but as yet there is no proof that it will be of benefit clinically in wound healing. EPIDERMAL GROWTH FACTOR EGF is a small molecule similar to TGF-α. EGF is produced by the platelet and found in high quantities in the early phase of wound healing. The active form has a molecular weight of 6200 Daltons. EGF is produced by the kidney, salivary glands, and lacrimal glands, and thus is found in high concentrations in urine, saliva, and tears. EGF promotes epidermal regeneration in pigs and corneal epithelialization in rabbits. It also increases the tensile strength of wounds in animals. EGF increases wound healing by stimulating the production of proteins such as fibronectin. Although EGF does not stimulate collagen production, it increases the number of fibroblasts in the wound. These cells produce collagen and improve the wound strength. EGF shares a receptor with TGF-α. EGF has been studied in a randomized trial of healing of skin graft donor sites (22). Donor sites treated with silver sulfadiazine containing EGF had an accelerated rate of epidermal regeneration as compared with patients treated with silver sulfadiazine alone. EGF reduced the healing time by 1.5 days. These results did not have clinical significance;

454 Steed however, this was the first trial to demonstrate a benefit from treatment with a single growth factor in human wounds. In another trial, EGF was used in an open label study with a crossover design in patients with chronic wounds (23). Patients with chronic wounds were treated with silver sulfadiazine. In those who did not heal, silver sulfadiazine containing EGF was then used. Improvement was noted in many of these patients. The results of these studies suggest that EGF may be of benefit in wound healing, although as yet, there is not enough data to confirm this. FIBROBLAST GROWTH FACTOR FGF is a group of heparin bound growth factors. There are two forms: acidic FGF (afgf) and basic FGF (bfgf). Both molecules have molecular weights of 15,000. There is a 50% amino acid homology among the two molecules. They are commonly bound to heparin or to heparan sulfate, which protects them from enzymatic degradation. FGF can be produced by fibroblasts, endothelial cells, smooth muscle cells, and chondrocytes. In addition to endothelial cells, FGFs can stimulate fibroblasts, keratinocytes, chondrocytes, and myoblasts. There are at least four different FGF receptors identified thus far. They appear to have a similar function. Both afgf and bfgf are found in the ECM in the bound form. Matrix degradation proteins then acts to release afgf or bfgf. Acidic FGF is similar to endothelial cell growth factor, whereas bfgf is similar to endothelial cell growth factor II. Both afgf and bfgf are similar to keratinocyte growth factor (KGF). These proteins are mitogens for cells of mesodermal and neuroectodermal origin. FGFs are potent mitogens for endothelial cells and function as angiogenesis factors by stimulating growth of new blood vessels through proliferation of capillary endothelial cells. To date, there are no clinical trials which have proven FGF to be of benefit in clinical wound healing. KERATINOCYTE GROWTH FACTOR KGF is closely related to the FGFs. It is a protein with a molecular weight of 2800 and has a significant amino acid homology with the FGFs. Although KGF is found only in fibroblasts, it stimulates keratinocytes, not fibroblasts. It may share a receptor with FGF. KGF-2 was used in a randomized prospective blinded trial of patients with venous stasis ulcers (24). There appeared to be a benefit from treatment with KGF-2. The role of KGF-2 in wound healing is still, however, unclear. INSULIN-LIKE GROWTH FACTOR IGFs, or somatomedins, are proteins that have a 50% amino acid homology with proinsulin and have insulin-like activity (25). There are two forms of this growth factor, IGF-1 and IGF-2. IGF-1 and IGF-2 are anabolic hormones that can stimulate the synthesis of glycogen and glycosaminoglycans. They can also increase the transport of glucose and amino acids across cell membranes. They increase collagen synthesis by fibroblasts. At this time, there are no clinical trials reported using IGFs. Both are secreted as large precursor molecules that are then cleaved to an active form. IGF-1 is identical to Somatomedin-C, whereas IGF-2 is similar to somatomedin. These growth factors are found in the liver, heart, lung, pancreas, brain, and muscle. IGF-2 is also synthesized by many different tissues but is particularly prominent during fetal development

Role of Growth Factors in Diabetic Foot Ulceration 455 and plays a significant role in fetal growth. IGF-1 and IGF-2 have separate receptors. The actions of pituitary growth hormone may be mediated through IGF-1. IGF-1 then causes cell division. IGF-1 is produced predominantly in the liver. It is found in high concentrations in platelets and is released into the wound when clotting occurs. Levels of IGF-1 and IGF-2 depend on many different factors, such as age, gender, nutritional status, and hormone level. Growth hormone is a regulator of IGF-1 and IGF-2 as are prolactin, thyroid hormone, and sex hormones. Elevated levels of somatomedins are found in patients with acromegaly. PLATELET RELEASATES In the first 2 days following injury, growth factors are produced and released by platelets. Thereafter, growth factor production is taken over by macrophages. Within the α-granules of the human platelet are multiple growth factors that are released when platelets are activated and degranulate. These include PDGF, TGF-β, FGF, EGF, platelet factor 4, platelet-derived angiogenesis factor, and β-thromboglobulin. A purified platelet releasate can be prepared by stimulating platelets to release the contents of their α-granules by using thrombin. Use of a platelet releasate in wound healing has theoretical advantages. The growth factors that are released are identical to and in the same proportion as those factors normally brought into the wound by the platelet. Preparation of a platelet releasate is simple and inexpensive because the platelets can be harvested from peripheral blood. Platelets readily release the contents of their α-granules when stimulated with thrombin. Growth factors are preserved in banked blood. Thus, large quantities of growth factors can be retrieved from the platelets of pooled human blood. There may, however, be disadvantages to using a platelet releasate. Not all growth factors promote wound healing. There is a signal for wound healing to stop. A platelet releasate might concentrate factors that heal the wound as well as those which signal the wound healing process to end. There is also the possibility of transmission of an infectious agent if the platelet releasate applied to the wound is from another individual. This risk could be reduced if the releasate were harvested from a single donor or from the patient himself. There has been considerable experience with the use of platelet releasates in wound healing. A preliminary report described the use of an autologous platelet releasate in six patients with chronic lower extremity ulcers from connective tissue diseases. A homologous platelet releasate was used to treat 11 patients with leg ulcers from diabetes and 8 patients with leg ulcers secondary to chronic venous insufficiency (26). No benefit was observed from using the platelet releasate; however, this study pointed out the importance of topical growth factor application only in the context of good wound care and in a narrowly defined group of patients. Growth factors cannot be expected to have a positive influence on wound healing unless they are applied in a comprehensive wound care program. The underlying etiology of these wounds such as venous hypertension, diabetic neuropathy, or ischemia must be addressed. In another trial, 49 patients with chronic wounds were treated with an autologous platelet releasate (27). There appeared to be a correlation with complete wound healing and initial wound size. This was the first clinical trial to suggest a benefit from a platelet releasate applied topically.

456 Steed A randomized trial of platelet releasate vs a platelet buffer was conducted in patients with ulcers secondary to diabetes, peripheral vascular disease, venous insufficiency, or vasculitis (28). This study suggested a benefit from the treatment of leg ulcers in these 32 patients from a topically applied growth factor preparation. The growth factor preparation was added to microcrystalline collagen, a potent stimulator of platelets. The exact contribution from the collagen to the healing of these wounds was not defined. Two other trials suggested a benefit from a platelet releasate. In one trial, patients were treated for 3 months with silver sulfadiazine (29). Only 3 of the 23 lower extremity wounds healed; however, when the platelet releasate was then applied, the remaining ulcers healed. Another study of 70 patients suggested a similar benefit from a platelet releasate (30). Despite evidence that platelet releasates are of benefit, another trial observed very different result. In a randomized prospective double-blind placebo controlled trial, topical platelet releasate was applied to the leg ulcers of 26 patients. The ulcers were secondary to diabetes, peripheral vascular disease, or chronic venous insufficiency (31). Wounds treated with the platelet releasate increased in size that is worsened, whereas wounds in the control group improved. This study suggested that a platelet releasate might be detrimental to wound healing. Thirteen patients with diabetic neurotrophic ulcers were enrolled in a randomized trial of a platelet releasate vs saline placebo (32). A benefit was seen in those treated with the platelet releasate. By 20 weeks of therapy, five of seven patients healed using the platelet releasate, whereas only two of six patients healed using the saline placebo. By 24 weeks of treatment, three additional patients in the control group healed, suggesting that the platelet releasate stimulated more rapid healing but did not result in a greater proportion of healed wounds. Although there is some evidence to suggest that they may be of benefit from a platelet releasate applied topically to lower extremity wounds, the inconsistency of the results as well as the concern about transmission of infectious agents in using a homologous preparation leaves their role in human wound healing undefined. Living skin equivalents (LSEs) are approved for use in humans. These LSEs are tissue grafts made from keratinocytes and dermal fibroblasts harvested from neonatal foreskin using tissue-engineering biotechnology (33). A graft cultured on a synthetic scaffold can be cryopreserved to prolong the viability of the cells and their metabolic activity (34). When placed on a wound, an LSE serves as a drug-delivery system providing matrix proteins and growth factors including PDGF, TGF-β, VEGF, and KGF to the wound (35). LSEs have no professional antigen-presenting cells, such as endothelial cells, leukocytes, Langerhans cells, or dermal dendritic cells; thus, they do not activate T cells and cause rejection. Fibroblasts and kerotinocytes do not express HLA class II antigens and thus do not activate unprimed allogenic T cells. The cells of the LSE are preprogrammed to die and are replaced by the patient s own cells in the healed wound. They stimulate epithelial cells to grow across the wound from the margin. LSEs have been shown to be of benefit in healing diabetic ulcers in certain cases (36). In summary, growth factors exert a powerful influence over wound healing, controlling the growth, differentiation, and metabolism of cells. There are only a few reports in which growth factors applied topically can exert a positive influence on wound repair;

Role of Growth Factors in Diabetic Foot Ulceration 457 however, there is no doubt that they control wound healing. It is likely that their actions will be defined and we will thus be able to control the wound environment to achieve a complete and durable wound healing. REFERENCES 1. Edington HE. Wound healing, in Basic Science Review for Surgeons (Simmons RL, Steed DL, eds.), WB Saunders, Philadelphia, 1992, pp. 41 55. 2. McGrath MH. Peptide growth factors and wound healing. Clin Plast Surg 1990;17:421 432. 3. Rothe MJ, Falanga V. Growth factors and wound healing. Clin Dermatol 1992;9:553 559. 4. Steed DL. Mediators of inflammation, in Basic Science Review for Surgeons (Simmons RL, Steed DL, eds.), WB Saunders, Philadelphia, 1992, pp. 12 29. 5. Grinnel F. Fibronectin and wound healing. Am J Dermatopathol 1982;4:185 192. 6. Folkman T, Lansburn M. Angiogenic factors. Science 1987;235:442 447. 7. Knighton DR, Hunt TK, Schewenstuhl A. Oxygen tension regulates the expression of angiogenesis factor by macrophages. Science 1983;221:1283 1290. 8. Knighton DR, Silver IA, Hunt TK. Regulation of wound healing angiogenesis: effect of oxygen gradients and inspired oxygen concentration. Surgery 1981;90:262 269. 9. Lynch SE, Nixon JC. Role of platelet-derived growth factor in wound healing: synergistic effects with growth factors. Proc Natl Acad Sci USA 1987;84:7696 7697. 10. Robson M, Phillips L. Platelet-derived growth factor BB for the treatment of chronic pressure ulcers. Lancet 1992;339:23 25. 11. Mustoe T, Cutler N. A phase II study to evaluate recombinant platelet-derived growth factor- BB in the treatment of stage 3 and 4 pressure ulcers. Arch Surg 1994;129:212 219. 12. Steed DL, Diabetic Ulcer Study Group: clinical evaluation of recombinant human plateletderived growth factor for the treatment of lower extremity diabetic ulcers. J Vasc Surg 1995;21:71 81. 13. Sporn MB, Robert AB. Transforming growth factor. JAMA 1989;262:938 941. 14. Assoian R, Komoriya A, Meyers C, Miller D, Sporn M. Transforming growth factor-beta in human platelets. Identification of a major storage site, purification, and characterization. J Biol Chem 1983;258:7155 7160. 15. O Kane S, Ferguson M. Transforming growth factor beta and wound healing. Int J Biochem Cell Biol 1997;29:63 78. 16. Ghahary A, Shen Y, Scott PG, Gong Y, Tredget E. Enhanced expression of mrna for transforming growth factor-beta, type I and type III procollagen in human post-burn hypertrophic scar tissues. J Lab Clin Med 1993;122:465 473. 17. Schmid P, Itin P, Cherry G, Bi C, Cox D. Enhanced expression of transforming growth factorbeta type I and type II receptors in wound granulation tissue and hypertrophic scar. Am J Pathol 1998;152:485 493. 18. Wang R, Ghahary A, Shen Q, Scott P, Roy K, Tredget E. Hypertrophic scar tissues and fibroblasts produce more transforming growth factor-beta 1 mrna and protein than normal skin and cells. Wound Repair Regen 2000;8:128 137. 19. Lee T, Chin G, Kim W, Chau D, Gittes G, Longaker M. Expression of transforming growth factor beta 1, 2, and 3 proteins in keloids. Ann Plast Surg 1999;43:179 184. 20. McPherson J, Pratt B, Steed D, Robson M. Healing of diabetic foot ulcers using transforming growth factor beta in collagen sponge. Abstracts of the meeting of the European Tissue Repair Society, 1999. 21. Howdieshell T, Callaway D, Webb W, et al. Antibody neutralization of vascular endothelial growth factor inhibits wound granulation tissue formation. J Surg Res 2001;96:173 182. 22. Brown GL, Curtsinger L, Nanney LB. Enhancement of wound healing by topical treatment with epidermal growth factor. N Engl J Med 1989;321:76 80.

458 Steed 23. Brown GL, Curtsinger L. Stimulation of healing of chronic wounds by epidermal growth factor. Plast Recontr Surg 1991;88:189 194. 24. Robson M, Steed D, Jensen J. Keratinocyte growth factor 2 in the treatment of venous leg ulcers. Abstracts of the World Congress of Wound Healing, 2000. 25. Spencer EM, Skover G, Hunt TK. Somatomedins: do they play a pivotal role in wound healing? in Growth Factors and Other Aspects of Wound Healing: Biological and Clinical Implications (Barbul A, Pines E, Caldwell M, eds.), Alan R Liss, New York, 1988. 26. Steed DL, Goslen B, Hambley R, et al. Clinical trials with purified platelet releasate, in Clinical and Experimental Approaches to Dermal and Epidermal Repair: Normal and Chronic Wounds (Barbul A, ed.), Alan R Liss, New York, 1990, pp. 103 113. 27. Knighton DR, Ciresi KF. Classifications and treatment of chronic nonhealing wounds. Ann Surg 1986;104:322 330. 28. Knighton DR, Ciresi K. Stimulation of repair in chronic, nonhealing, cutaneous ulcers using platelet-derived wound healing formula. Surg Gynecol Obstet 1990;170:56 60. 29. Atri SC, Misra J. Use of homologous platelet factors in achieving total healing of recalcitrant skin ulcers. Surgery 1990;108:508 512. 30. Holloway GA, Steed DL, DeMarco MJ, et al. A randomized controlled dose response trial of activated platelet supernatant topical CT-102 (APST) in chronic nonhealing wounds in patients with diabetes mellitus. Wounds 1993;5:198 206. 31. Krupski WC, Reilly LM. A prospective randomized trial of autologous platelet-derived wound healing factors for treatment of chronic nonhealing wounds: a preliminary report. J Vasc Surg 1991;14:526 532. 32. Steed DL, Goslen JB, Holloway GA. CT-102 activated platelet supernatant, topical versus placebo: a randomized prospective double blind trial in healing of chronic diabetic foot ulcers. Diabetes Care 1992;15:1598 1604. 33. Rennekampff H, Kiessig V, Hansbrough J. Current concepts in the development of cultured skin replacements. J Surg Res 1996;62:288 295. 34. Hansbrough J, Franco E. Skin replacements. Clin Plast Surg 1998;25:407 423. 35. Yannas I. Studies on the biological activity of the dermal regeneration template. Wound Repair Regen 1998;6:518 524. 36. Pollak R, Edington H, Jensen J. A human dermal replacement for the treatment of diabetic foot ulcers. Wounds 1997;9:175 183.