The Diabetic Foot

At a symposium on wound healing, Harold Brem (New York, NY) noted that nearly 100,000 lower-limb amputations are performed annually on individuals with diabetes in the U.S. (1). Lower-extremity ulcers occur in 4–10% of people with diabetes, with a lifetime risk that may be as high as 25% (2). Brem reviewed an approach to diabetic foot wounds based on debridement, offloading, and moist wound healing, and suggested that such an approach to diabetic foot ulcers would lead to “a dramatic decrease in amputations.” He pointed out that often the diabetic foot wound “does not look bad,” contributing to inadequate treatment, and asked why “these good-looking wounds” do not heal. Two factors are underlying osteomyelitis, often with relatively little evidence of inflammation, and arterial insufficiency. Brem reviewed data from the treatment of >250 patients, approximately one-fifth of whom required amputation, with over half ischemic and over half having underlying osteomyelitis.

Debridement needs to include not only the wound and the surrounding callus, but also the “physiologically impaired area.” Abnormal keratinocyte products may be demonstrated outside the area of callus (3), suggesting that typical debridement is insufficient to reach the actual “healing edge” (4), the area with molecular markers of appropriate wound healing. An important tool Brem uses to track the progress of wound healing is the “wound electronic medical record,” which includes objective measures of wound size, ischemia, and infection, as well general factors pertaining to glycemic control and diabetic complications. Although there is controversy as to whether risk increases with aging, Brem showed evidence that the combination of age and diabetes leads to impaired healing (5). Addressing the question of ischemia, Brem commented that “the most important thing … is to immediately screen with an ankle-brachial index.” It is important that vascular intervention be performed before amputation, regardless of the planned level.

Growth factor treatment may be a promising approach to diabetic foot wounds. Vascular endothelial growth factor (VEGF) treatment in a mouse model with an adenovirus vector significantly decreased the time to wound closure. In addition, it led to increased epithelialization as well as to keratinocyte and fibroblast migration into the wound, increasing collagen deposition. This treatment is only effective, however, when sufficient wound debridement is done, as was earlier shown with platelet-derived growth factor (PDGF). Such treatment gains additional rationale from the impairment in release of both VEGF and PDGF in diabetes, particularly in cells located in the central area of the wound (6). Brem suggested pathologic evaluation of debrided tissue to make sure that debridement includes tissue beyond the area of hyperkeratotic material to ensure that adequate healing will occur, an approach that will lead to the presence of adequately responsive keratinocytes and fibroblasts and the inward migration of macrophages. Furthermore, cells from the nonhealing portion of the wound contain pathogenic molecules that delay healing, so that full removal is important.

Brem concluded that such a protocol, as well as offloading and use of evidence-based antibiotics if infection is documented (with bone scan or magnetic resonance imaging), with growth factors and cell therapy, and with surgical debridement based on histology, is feasible and will greatly improve outcome. Future developments may include assays for molecular markers of inhibition of wound healing to improve approaches to debridement, as well as new treatments based on the biology of diabetic foot ulcers, perhaps involving sustained delivery of growth factors and angiogenic factors.

Loretta Vileikyte (Manchester, U.K.) discussed the interrelationship between stress and wound healing, reminding the audience that “folk wisdom has long held” that there is a relationship between stress and susceptibility to infection. Stress increases susceptibility to infection and reactivates latent viral infection. She stated that the strongest evidence of this phenomenon is in the field of wound healing, according to a review of studies in human acute experimentally induced wounds, operative wounds, and chronic ulcers.

The term “stress” is not a simple one and may be conceptualized in a variety of fashions. From an environmental perspective, stress may be caused by documented adverse life events objectively associated with adaptive demands, such as bereavement and loss of work. From a psychological perspective, stress is based on individuals’ subjective evaluations and emotional responses to demands. Finally, from a biological perspective, the degree of stress may be quantitated from measures of activation of physiological systems, such as elevated circulating levels of cortisol and catecholamines, with all these forms of stress closely linked. Environmental stimuli lead to negative emotions and behavior, with subsequent biological response. The physiology of stress involves the hypothalamic/pituitary/adrenal axis, with cortisol-releasing hormone stimulating adrenocorticotropic hormone, leading to adrenal glucocorticoid production, while the sympathetic axis is activated via innervation of adrenal medulla. Glucocorticods decrease insulin-like growth factor (IGF)-1, leading to impaired fibroblast function, and cortisol may lead to cytokine-induced tissue damage. Depression and anxiety increase circulating cortisol levels, as well as levels of proinflammatory cytokines, with additional hormonal and immunological effects. In animal models, restraint stress is associated with higher circulating corticosteroid levels and with a 27% reduction in wound healing (7).

In a study of 47 persons following inguinal hernia surgery, the “perceived stress scale” and worry scores predicted a lowering of wound fluid levels of matrix metalloproteinase and interleukin (IL)-1, suggesting that psychological stress impairs the inflammatory responses and matrix degradation processes normally seen following a surgical intervention (8). Stress also can lead to decreased sleep, poor nutrition, less exercise, abuse of alcohol and drugs, and reduced medication compliance. All of these factors may play roles in wound healing, with Vileikyte citing a study of 28 older adults undergoing dermal punch biopsy. Those engaging in regular exercise showed wound healing in a mean of 29 days, while sedentary individuals averaged 39 days for healing. She speculated that beneficial effects of social support may involve oxytocin, a neuropeptide involved in social behaviors that also promotes wound healing. Slow wound healing is seen in persons with lower levels of hopefulness. In a study of caregivers of individuals with dementia, who presumably are under increased levels of stress, poor healing of dermal punch biopsy was demonstrated, with production of IL-1β peripheral message lower in their peripheral blood leukocytes (9). A study of stress induced by a pending examination in dental students showed a 40% reduction in healing of hard palate biopsies, again with reduced leukocyte IL-1β production (10).

Skin blisters from persons with higher perceived stress have lower IL-1α and IL-8, while salivary cortisol levels are lower in such individuals (11). Similarly, although married people in general have lower mortality and morbidity, marital stress is also associated with reduction in blister wound healing with reduced wound cytokine levels, but increased circulating IL-6 and tumor necrosis factor (TNF)-α are found in more hostile couples (12). Depression and postsurgical pain are similarly associated with reduced healing of punch biopsy wounds (13).

Clinical studies have shown association between stress and outcome of wound healing. In a study of patients with chronic venous ulcer, healing showed inverse association with measures of anxiety and depression, with patients in the top half of these scores four times more likely to have delayed healing than those in the lower half (14). Thus, it is important to consider the relationship between stress and healing of chronic wounds in diabetic lower-extremity ulcers. Diabetic neuropathy is strongly associated with anxiety and depression, presumably mediated by factors such as patients’ perceptions of the unpredictability of neuropathy symptoms, perceived inability to control outcome, and restrictions in activity and changes in social self-perception caused by increasing degrees of disability (15). Furthermore, the presence of depression is associated with a tripling of mortality among individuals with diabetic foot ulcers (16). These associations have many potential confounders, including the effects of frequent surgery for debridement, ischemia, and amputation, and the need for high levels of adherence to off-loading. Nonetheless, the concept that a physiological stress response, mediated by abnormalities of systemic and local cytokines and stress hormones, mediates effects of adverse emotions on health appears valid, with increasing evidence that stress downregulates the endocrine and immune systems needed for effective wound healing.

Robert Kirsner (Miami, FL) discussed a number of new approaches to wound healing. Many guidelines and algorithms have been proposed for leg ulcers; these have included recommendations for evaluating for ischemia and osteomyelitis, performing debridement, moist wound healing, and off-loading (17). Kirsner noted, however, that although offloading is effective, it is crucial that the patient use it, explaining the benefit of nonremovable casts is “to make it impenetrable to the patient,” but most studies suggest that patients only wear the devices approximately one-quarter of the time. Seemingly similar approaches, such as those using the Una boot, are “good for compression, [but do not] provide off-loading.” Given the difficulties in compliance, it may then not be surprising that patient outcomes are relatively poor with approaches following existing wound care guidelines. In a systematic review published in 1999 of the control groups of nine randomized controlled trials, at 12 weeks, complete closure had occurred in 24%, and at 20 weeks, in only 31% treated in this fashion (18). Further analysis of this database of 586 subjects, all of whose wound care included debridement and off-loading, showed that although smaller size and shorter duration of ulcers were associated with greater likelihood of healing, general clinical factors were not associated with outcome, with neither age nor glycosylated hemoglobin predicting likelihood of wound healing (19).

Kirsner cited a more recent study based on a database of >20,000 nonischemic wounds, showing only 30–45% of diabetic foot ulcers healing over a 6-month period. Patient compliance, stress, inconvenience, and clinician knowledge are all important factors, and a number of approaches to adjunctive treatment have been explored. Platelet-derived growth factor (PDGF) has been available as Regranex since 1997, and two artificial skin materials are commercially available, Apligraf (since 2000) and Dermagraf (since 2001). While none of these agents lead to dramatically better wound healing, all do improve outcome to some degree (20–22).

Kirsner discussed a number of promising future approaches. The environment of the chronic wound contains unresponsive and senescent cells, bacteria, deficient and/or unavailable growth factors, and an inflammatory and proteolytic environment. The latter findings are characteristic of an early phase of wound healing, seen in normal acute wounds immediately after the phase of coagulation. Normally, however, inflammation is followed by a phase of migration and proliferation of cellular elements involved in tissue repair, after which remodeling occurs to complete the process. Chronic wounds are halted in the inflammatory phase, with proteases, particularly matrix metalloproteinases, appearing to be particularly important in interfering with ongoing healing. Part of the benefit of debridement is in allowing a wound to move past the inflammatory phase.

A number of agents are being developed for the treatment of chronic ulcers. Topical application of doxycycline may have benefit. A thrombin peptide, chrysalin, has shown some evidence of improvement in wound healing (23). Topical adenosine agonists may improve wound healing, as neutrophil adenosine A2 receptors inhibit phagocytosis and secretion of inflammatory molecules and reduce oxidative stress, with these agents potentially increasing IL-10 secretion by monocytes, leading to anti-inflammatory effects and enhanced platelet aggregation. MRE0094, a gel developed by King Pharmaceuticals for topical use, has such an effect and is in phase II studies. Human lactoferrin is a nonheme iron-binding protein found in neutrophils, in serum, and in exocrine secretions. Keratinocyte secretion of TNF-α and IL-1b has a proinflammatory effect and is inhibited by lactoferrin, with recombinant lactoferrin appearing to be effective in promoting wound healing in animal models. However, in clinical trials of a synthetic form of this agent developed by Agennix, only a nonsignificant trend to wound healing improvement was reported (24). TNF-α expression is increased in nonhealing chronic wounds and may be important in diabetic wounds. An approach with anti–TNF-α treatment has been described to improve experimental wounds in ob/ob mice (25). Local growth factor treatment is in part impaired by the inflammatory environment of the wound, with investigations being carried out using a gene-activated matrix approach based on a PDGF-expressing adenovirus treatment applied locally. Growth factor delivery by special preparations of Apligraf or Dermagraf, or by using keratinocyte- or fibroblast-coated microspheres, ia also being studied, with animal models suggesting that the approach stimulates fibroblast migration, collagen production, and angiogenesis. A folk remedy derived from the bark and fruit of horse chestnut may also play a role in promoting wound healing.

There is a well-recognized relationship between sensory neuropathy and loss of protective sensation, but Kirsner pointed out that neuropathy also results in loss of cytokines normally delivered by peripheral nerves, which contribute to wound healing. The neurotransmitter substance P also may favorably promote wound healing. Fetuses heal differently from adults, without scarring, an effect possibly involving neural stimulation, as denervation of fetal lamb wounds slows healing rates (26). There has been research suggesting benefit of topical nerve growth factor treatment on wound healing (27), and epidermal growth factor (EGF) also has been shown to improve healing (28), a potential mechanism of the long-used burn treatment silvadene, which appears to increase local EGF levels.

Another area of potential benefit in wound healing may involve stimulation of angiogenesis. It should be noted that a key mechanism suggested for benefit of hyperbaric oxygen, to be discussed in next month's column, is in increasing levels of angiogenesis. A topical VEGF treatment being developed by Genentech, Telbermin, showed a trend to improved healing in a phase I study. Angiogenesis also may be stimulated by acetylcholine via the nicotinic receptor, which induces endothelial cell proliferation and migration and mediates angiogenesis. Inhibition of this receptor is being studied for cancer treatment, and topical nicotine, which stimulates the receptor, promotes angiogenesis in tumor models (29) and is being used in wound healing studies, recalling the folk remedy of tobacco poultices. A nicotine-containing gel, applied in a lower concentration than used for smoking cessation, stimulates angiogenesis, cellular proliferation, and migration and may have anti-microbial effects, with studies in the db/db mouse showing improved healing, although Kirsner reminded the audience (with tongue in cheek) that these experiments should not be taken as a rationale to encourage persons with foot wounds to begin or continue cigarette smoking!

In a relevant study presented at the ADA meeting, Kipnes et al. (abstract 8) treated 24 diabetic individuals with peripheral neuropathy by injecting plasmid DNA expressing a zinc finger transcription factor for VEGF in the legs, showing improvement in motor nerve conduction velocity, in vibration sensation perception, and in a clinical neuropathy symptom score. (Abstract numbers refer to the American Diabetes Association Scientific Sessions, Diabetes 56 [Suppl. 1], 2007.)

Taehee Kwak (Boston, MA) discussed stem cells and wound healing. Stem cell treatment of chronic wounds may be of benefit because stem cells present in wounds may have been altered in such a fashion as to become less effective. Bone marrow–derived stem cells may then be suitable for treatment, once effective methods of delivery become available. The mechanism of action of these cells may involve their actual engraftment or may involve factors they release that lead to short-term stimulation of healing, presumably in a fashion additive to that of other treatments. Upcoming studies will compare autologous with allogenic bone marrow, and will endeavor to develop optimal delivery vehicles, perhaps with co-administration of other stem cell types.

In the diabetic ulcer, cells from debrided callus and other infected materials are phenotypically altered. When these cells are grown in vitro, fibroblasts and myofibroblasts are formed with abnormal morphology, low proliferation, and abnormal response to growth factors such as PDGF and transforming growth factor (TGF)-β, with evidence of downregulation of TGF-β receptors and decreased downstream cellular mitogenic activity (30). Two stem cell lines are present in bone marrow, stromal and hematopoietic. Kwak reviewed preliminary human studies suggesting that bone marrow–derived cultured cells applied to a chronic wound under an occlusive dressing improve healing. A human study used cultured autologous bone marrow cells showing cell markers for mesenchymal stem cells (MSCs) administered in a fibrin spray delivery system. Potentially beneficial effects on wounds included generation of new elastic fibers (31). In mouse wound models, treatment with autologous MSCs is similarly associated with improved healing (32). Another approach, with studies in progress, involves use of bioengineered skin to guide autologous MSCs. Thus, stem cells may offer a promising approach to treatment of diabetic wounds, with the fibrin gel or spray delivery system appearing to represent a promising approach, although a great deal of work will be needed to determine appropriate clinical approaches using these methods.

Lawrence Lavery (San Antonio, TX) presented the annual Roger Pecoraro Lecture on the topic of the prevention of diabetic foot complications, characterizing this as “a missed opportunity in the disease management world.” A 1998 survey carried out by the Rand Corporation showed that only half of the U.S. population received recommended preventive care, even among persons with known underlying chronic illness (33). Traditional disease management programs have focused on managing outliers, rather than on saving money by improving quality, a goal that Lavery suggested might be particularly possible with the diabetic foot. In treatment of individuals with diabetes, such disease management programs typically address cardiovascular and renal disease and diabetic retinopathy, rather than foot complications. Yet, the diabetic foot is an area in which highly cost-effective treatment is possible, and this area represents a good opportunity for preventive care of a group of persons for whom expensive and recurrent complications occur frequently. Given these considerations, Lavery said, it should be considered alarming that “the foot isn’t even on the radar screen for most people,” with the use of correct interventions in appropriate target populations important for physicians caring for diabetic patients to “do what we already say we are doing,” potentially reducing adverse outcome with actual cost savings.

The incidence of foot wounds is 2–7% per year among individuals with diabetes, with the majority ending in amputation at some level, and with 20–80% rates of recurrence. Reviewing risk factors for amputation, Lavery noted the higher incidence among diabetic persons from ethnic minorities in the U.S., and that men have twice as great an incidence of amputations as women. If sensory loss and repetitive trauma, limited joint mobility, poor healing, and ischemia lead to ulcers and amputations, intervention at the level of individuals with these risk factors can reduce these complications. Certain risk factors help in stratification, particularly previous foot pathology and peripheral vascular disease, as well as, to a lesser extent, the identification of persons with sensory loss. Using such an approach, it is possible to identify the 20% of diabetic individuals with 70% of ulcers and 90% of amputations. In agreement with Brem's comments, Lavery suggested that measurement of the ankle-brachial index to screen for arterial insufficiency may be particularly useful as a screening tool, perhaps more useful than sensory testing. Approaches such as foot care education, provision of appropriate footwear, and regular follow-up are of great importance for those at high risk for diabetic foot complications. Footwear may reduce amputation by half when it is used, and Lavery stressed the importance of appropriate self-care, noting that often a relative must be involved, as many patients cannot themselves inspect their feet because of obesity.

In his group's study, however, only two-thirds of at-risk individuals had had foot care education, and only one-third had received appropriate footwear, with those persons obtaining such treatment on average 2–7 years previously. The intervention of lengthening the Achilles tendon for individuals with equinus deformity is highly effective in decreasing plantar forefoot pressure and in optimizing the healing of foot ulcers, yet relatively few persons who might benefit from such treatment receive it.

Lavery reviewed his studies with infrared thermometers for home temperature monitoring to identify areas of injury before development of recognizable signs and symptoms (34,35). In three randomized controlled trials of >300 high-risk individuals without arterial insufficiency, a 3- to 10-fold reduction in ulceration occurred, despite the control group receiving appropriate footwear and education, showing that visual inspection and standard care are in themselves insufficient to prevent foot ulceration. He gave some anecdotal cases of individuals using infrared thermometer devices that suggest that after an injury, there is opportunity for prevention of ulceration for up to 2–3 weeks. New, simpler devices are being developed that allow at-risk persons to stand on a measuring surface for assessment of foot temperature. With these devices in his studies, for every five individuals treated, a foot wound could be prevented over a period of 6–12 months. Similar benefits can be demonstrated for footwear; Lavery stressed the need to “take away the shoe as a source of pathology” simply by making sure it fits properly. “We are moving to a more scientific approach to the design of shoes and insoles,” he said. Similar benefits, he stated, are seen for Achilles tendon lengthening. Lavery contrasted these simple treatments with the use of angiotensin-converting enzyme inhibitors and statins, for which 20–40 persons need to be treated over 2–5 years to prevent one event. “Perhaps,” he concluded, “the foot can be the driver for initial savings and improvement in quality.”

David Armstrong (Tucson, AZ) continued the discussion with a summary of “common sense approaches to diabetic footwear and management.” He addressed questions of approaches to wound debridement. There is enormous variability in the rate of amputation from practitioner to practitioner, and from region to region, suggesting that we have not yet found a way of reproducibly delivering optimal care at many different sites. Most amputations are not performed because of ischemia per se, but rather because of infection, with the degree of vascular disease determining the level at which amputation is required. In nonischemic wounds, debridement and off-loading, as well as use of dressings and other local treatments, are important.

Armstrong reviewed data from a study of recombinant human platelet-derived growth factor for the treatment of lower extremity diabetic ulcers, comparing centers which debride more versus less, showing that the degree of healing was related to the intensity of wound debridement (36), so that this approach “must be helping.” Armstrong further analyzed data from his center that suggested the benefit of debridement and reviewed data showing that diabetic individuals with foot wounds treated with more intensive debridement have improvement in healing, with the cost of debridement relatively low, given its benefit. Debridement may involve cutting, or ultrasound, or whirlpool, or even a water-pick device. Armstrong reviewed an approach to debridement by application of sterile green blow fly larvae, which secrete digestive enzymes and appear to effectively initiate granulation, perhaps based on mechanical stimulation or factors in larval secretions. Larval debridement decreases the surface area of necrotic tissue and increases the surface area of granulation tissue and may be cost-effective (37). In a study of 13 persons with methicillin-resistant Staphylococcus aureus, larval application completely eradicated infection in 12 of the subjects (38). Armstrong concluded that aggressive debridement, followed by effective offloading, followed by dressings that do not require frequent changes, may be the bases of development of a reliable approach to the treatment of the diabetic foot wound.

A Web site with resources on the use of larvae for debridement therapy is http://www.ucihs.uci.edu/som/pathology/sherman/home_pg.htm. Also, the 2007 update of the International Consensus on the Diabetic Foot and Practical Guidelines on the Management and Prevention of the Diabetic Foot is available as an interactive DVD at www.idf.org/bookshop. In a study presented at the ADA meeting, Game et al. (abstract 259) questioned whether early surgery is required for diabetic patients with foot ulcers and evidence of osteomyelitis, reporting a series of 147 such patients. Approximately 80% of those with infection of the forefoot and 60% of those with mid- to hindfoot infection resolved with broad-spectrum antibiotics and without surgery.