INTRODUCTION Leg ulceration is a common disorder, though the incidence and prevalence have not been well established. It is estimated that 0.12 to 0.19 percent of Western populations have leg ulcers, and in people aged 65 or older the prevalence of venous leg ulcers is estimated at 1.0 to 3.3 percent. Seventy to 81 percent of leg ulcers are caused by venous disease, and arterial disease accounts for another 10 to 25 percent, which may coexist with venous disease. Coexisting rheumatoid disease occurs in nine percent of wound patients, whereas diabetes mellitus is present in 5 to 12 percent of patients. Less commonly, trauma, pressure, inflammatory disorders, and infectious agents also are causes of leg ulcers. The overlap of various causes, as well as coexisting disease, occurs because these conditions are not mutually exclusive.[1–7] Because leg ulceration is a condition characterized by chronicity and relapse, it gives rise to massive healthcare expenditure. In the European countries, the care of patients with venous leg ulcers consumes 1 to 2 percent of the overall healthcare resources.[8] A chronic wound is predisposed to secondary infections, which can lead to amputations, which in turn are associated with increased mortality. Each person with a chronic wound suffers from pain and discomfort, and a chronic wound entails restrictions in a person’s everyday life and social activities.[9] Since most leg ulcers are of venous origin, they can be healed with compression therapy alone or combined with local treatments. In recent years, the concept of a clean, moist environment has been widely accepted in the treatment of leg ulcers. In some cases, however, a moist environment and compression therapy are not sufficient for the ulcer to heal, and for these hard-to-heal ulcers, the development of tissue-engineered products can offer new options for treatment.[10,11] A tissue-engineered human dermal substitute (HDS) (Dermagraft®, Smith & Nephew Inc., Largo, Florida) is designed to replace the dermis and to provide essential stimulatory growth factors for wound healing. It contains living human fibroblasts obtained from neonatal foreskins seeded onto a bioabsorbable polyglactin mesh.[12,13] The HDS pieces are stored frozen at -70°C until used. Although the precise mechanism of action of tissue-engineered dermis is not completely understood, it has been shown to provide elements to the wound bed that are believed to be important in the repair process. HDS provides live, nonsenescent fibroblasts capable of colonizing the wound bed and persisting in situ for several weeks. The fibroblasts are capable of secreting a number of cytokines and growth factors, including platelet-derived growth factor, insulin-like growth factors I and II, heparin-binding epidermal growth factor, vascular endothelial growth factor, transforming growth factors a and b, and keratinocyte growth factor. Growth factors are known to stimulate fibroblasts, granulation tissue, matrix deposition, angiogenesis, and skin cell maturation. The fibroblasts also produce matrix proteins like collagen types I and III, fibronectin, and tenascin, as well as glycosaminoglycans, which bind growth factors and enhance their activity.[14,15] The overall safety and lack of rejection reactions combined with the efficacy[16,17] encourages the use of HDS in addition to good wound care practices. In the present paper, the authors describe their experience in the treatment of leg ulcers of various origins with HDS. Patients and Methods This was an open, noncomparative, retrospective analysis to assess the use of HDS in the treatment of leg ulcers of various origins when conservative treatment with compression and various local treatments had proved ineffective. The patients were treated at the clinics of dermatology in Tampere University Hospital, Tampere, Finland, and in the Jyväskylä Central Hospital, Jyväskylä, Finland. All patients had chronic (>3 months of duration), full-thickness (>1 cm2) ulcers, and previous conservative methods had proven ineffective. Demographic parameters, associated diseases, and medical and ulcer histories were assessed. The ulcers were recorded by Polaroid or digital photography, and the sizes of the ulcers were measured in square centimeters. The etiology of the ulcers was established together by clinical examination, through the use of a hand-held Doppler instrument, and through skin biopsies. Although the etiology of a chronic leg ulcer is often multifactorial, only one main etiologic factor for each patient’s leg ulcer was selected when possible. Patients using medications known to interfere with healing (e.g., corticosteroids, immunosuppressives, or cytotoxic agents) were not excluded because it is known that chronic inflammatory ulcers, such as those caused by vasculitis, rheumatoid arthritis, or pyoderma gangrenosum, may require these agents to bring the disease activity under control and thus provide optimal circumstances for wound healing. The ulcer was considered venous (28 patients) if there was a typical clinical picture of venous insufficiency detected with venography, edema of the leg, and normal arterial circulation recorded by phlebography or detected by examinations conducted by vascular surgeons. The diabetic ulcers (24 patients) were determined when there was a patient history of insulin- or noninsulin-dependent diabetes mellitus and presence of a normal or a high ankle-brachial pressure index (ABPI) without severe venous or arterial insufficiency. All diabetic ulcers were neuropathic. The ulcers were considered rheumatic in patients who suffered from rheumatoid arthritis (14 patients). In this group, active vasculitis was excluded with skin biopsies and no severe venous insufficiency or arterial disease was detected. Vasculitic ulcers (8 patients) were diagnosed through histopathology and immunohistochemistry of skin biopsies. Postradiation ulcers, ulcers appearing after skin biopsies taken to diagnose skin disease, and self-induced ulcers are examples of post-traumatic ulcers (8 patients). There were nine other ulcers, described here as a separate group (Other) with various incidental causes, such as ulcers caused by pyoderma gangrenosum (3 patients), necrobiosis lipoidica (3 patients), or scleroderma (1 patient). The authors used HDS on 13 ulcers classified as arterial ulcers when other etiologies could have been ruled out and when ABPI was 0.9 or less, the lowest being 0.5. In 10 cases, it was impossible to determine the major cause for the ulcer (e.g., one patient with diabetes developed an ulcer post-traumatically and had signs of arterial compromise), and the data of these patients is presented as a group of mixed ulcers. The ulcers studied were pretreated until they were free of necrotic material and infection and were suitable for HDS treatment (no exposed tendon, bone, or joint, and no tunnels or sinus tracts that could not be debrided). The frozen HDS pieces were defrosted in a water bath and then applied to the ulcers. The HDS implant was reapplied weekly. A nonadherent primary dressing (Tegapore®, 3M Inc., St. Paul, Minnesota) was taped over HDS followed by an absorbing secondary polyurethane foam dressing (Allevyn®, Smith & Nephew Inc., Largo, Florida). All the patients with venous leg ulcers were instructed to wear compression bandages or medical stockings, as were all the other patients when there was no evidence of severe arterial compromise. If the drainage was excessive during the week, the patients were instructed to change only the outer dressing. Wound closure was defined either as full epithelization with no drainage or as a reduction of wound area to 1mm2. The patients were assessed weekly until healing or withdrawal. The patients were withdrawn from the treatment if there was presence of uncontrolled wound infection, lack of patient adherence to the treatment plan, or lack of treatment efficacy. Follow-up phone calls were made approximately two years after the end of the treatment (range 1–3 years) to ensure the long-term treatment effect. The statistical analysis methods used were the Chi-square test and the logistic regression tests for the bivariate comparisions. Statistical significance was determined at the p=5cm2 (95% CI 0.908–4.285, p=0.0863). If the wound had existed less than one year compared to one year or more, the healing rate was 1.3 times faster, but these results were not statistically significant. The results were slightly better among women compared to men; the reduction in size was 84 percent compared to 70 percent, respectively, with a p value of 0.081. Ulcers among patients less than 60 years old were reduced in size 83% compared to 67% in patients 60 years old and older (p=0.081). The immunosuppressive medications had no effect on the healing of the ulcers. The reduction in the size of the ulcers was 83 percent if no immunosuppressants were taken compared to 76 percent for the patients with immunosuppressive medications. This was not statistically significant. Four patients died during the follow up; two of them were older persons with diabetes and two were older rheumatic persons. The amputation rate was three percent. Two of the patients with diabetes developed new leg ulcers, and one of the patients with diabetes had pyoderma gangrenosum ulcers and also signs of arterial disease. Figure 1 shows that rheumatic ulcers diminished 82 percent in size when compared with the wound size at the beginning of the treatment versus at the end of the treatment. Table 1 shows that 93 percent of the rheumatic ulcers were totally healed and 79 percent remained healed in the follow up. Figure 2 shows the numbers of patients by wound etiology. The arterial ulcers diminished in size 65 percent. Only 31 percent of these ulcers were totally healed, and at follow up, only 15 percent were healed. The diabetic wounds were 64-percent smaller at the end of the study. Figure 1 also shows that the standard deviation in this group was very large; only 15 percent were healed as shown also in Figure 3 and Table 1. In the group of post-traumatic ulcers, the mean wound size reduction was found to be 62 percent and the standard deviation was large. The percentage of healed wounds was 62 percent; during the follow-up period, this ratio fell to 50 percent. Thirty-eight percent of the vasculitic ulcers were healed with the mean wound size reduction of 58 percent. In this group, there was also a significant deviation in wound size reduction. Figure 1 also shows that venous ulcers were 52-percent smaller at the end of the study and that the standard deviation was largest in this group. On the other hand, as shown in Table 1, 61 percent of the wounds were totally healed, and by the follow up, this percentage was 46. In the Mixed and Other Ulcers groups, 50 percent and 56 percent of the wounds were healed, respectively. On the other hand, as shown in Table 1 and Figure 3, 61 percent of the wounds were totally healed, and by follow up, this percentage was 46. Discussion This study provides pilot experience about using HDS in hard-to-heal ulcers of various origins. The authors found that important predictors of ulcer healing were ulcer size and ulcer duration. Smaller ulcers of any origin seemed to respond to treatment better than larger ones as also described by Margolis, et al.,[18] in a study for diabetic foot ulcers. On the other hand, Gentzkow, et al.,[13] used HDS to treat diabetic foot ulcers, and according to their results, the ulcer size did not predict which ulcers healed. It has been found that only 13 percent of the venous ulcers experience healing with compression therapy alone if the wound is larger than 5cm2 and is present for more than six months,[19] and it is then justified to try various treatment options for these hard-to-heal venous ulcers. The venous ulcers enrolled in this study had not responded to conservative treatment with compression bandages and various local treatments. However, 61 percent were healed when treated with the HDS, and 46 percent remained healed at follow up. An earlier study on the use of HDS in treating venous leg ulcers revealed that 38 percent of the wounds healed compared to only 15 percent in the multilayer compression bandage control group.[20] A Finnish study showed promising results when venous leg ulcers were treated with a living skin equivalent (LSE) (Apligraf®, Organogenesis Inc., Canton, Massachusetts). In that study, 60 percent of the wounds were healed and some reduction in size was seen in all of the study ulcers (n=11).[21] According to Falanga, et al.,[22] it has also been found that significantly more patients healed in six months when treated with LSE plus compression therapy than with compression therapy alone (63% vs. 49%; p=0.02). LSE was particularly effective in difficult-to-heal venous ulcers, ulcers of more than six months duration, and larger and deeper ulcers. The present results suggest that the treatment of venous ulcers with HDS may be similar to the reported studies on LSE; however, no comparative trials have yet been conducted. Unfortunately, no comparative trials between the treatment of venous leg ulcers with surgical treatment and cultured human skin derivatives have been carried out. Surgical treatment successfully has been used for the treatment of chronic venous leg ulcers, with healing rates up to 78 percent.[23] Pinch grafting techniques also offer useful options for smaller venous wounds, and Öien, et al.,[24] have reported a 45-percent healing rate for venous ulcers. Most arterial leg ulcers do not meet the criteria of chronic critical limb ischemia, but they do not heal with conservative methods either.[25] The present results show that arterial ulcers responded least to the treatment with HDS. Still, 31 percent of the patients healed, and during the follow up this ratio was 15 percent. This result was better than expected because no surgical interventions were made during this study. Hafner, et al.,[25] found that ABPI is a fairly inaccurate test to distinguish the patients who require revascularization from those who will heal conservatively. It is justified to start with a conservative treatment when the arterial ulcer is small and with little pain. If no or weak response is achieved, the patient should be evaluated for skin grafting or treatment with skin replacement products along with revascularization. The need for new therapies for diabetic ulcers can be justified when the frequency of nonhealing is 67 percent for diabetic foot ulcers after 20 weeks of treatment with standard care as stated by Kantor and Margolis.[26] In the present study, the ulcers of diabetic origin were healed in 63 percent of the cases, and during the follow up, this ratio was 54 percent. These results can be considered very good, since in a prospective, randomized, multicenter study by Gentzkow, et al.,[13] only 50 percent of the HDS-treated and eight percent of the control ulcers healed completely (p=0.03). In that study, after a mean of 14 months of follow up no recurrence was observed in the HDS-treated ulcers, which differs from the results with 13-percent recurrence. The present results show that HDS could be especially beneficial in the treatment of chronic rheumatic ulcers, which, when treated with HDS, healed with rates up to 93 percent during the treatment period; seventy-nine percent were still healed at the follow up. It is thought that rheumatic ulcers often heal poorly due to skin fragility as a side effect of poor nutrition, corticosteroids, nonsteroidal anti-inflammatory drugs (NSAIDs), or immunosuppressive therapy, together with foot deformity. These patients are often less suitable for surgical operations, and HDS may provide essential growth factors for these wounds to heal. In ulcers of rheumatic origin, signs of venous insufficiency also can be seen due to calf pump failure resulting from immobility, reduced ankle movements, and possibly venous thromboses.[27] The compression therapy given may explain some but not all of the success achieved in the treatment of rheumatic ulcers with HDS. The treatment of vasculitic ulcers still remains a problem. HDS showed efficacy in a few cases in which the inflammatory process had been suppressed, but it is worth trying together with immunosuppressive medications or as an alternative for pinch grafting in very complex cases.[28,29] In the group of the mixed ulcers, the healing rate was 50 percent, which can be considered excellent because these ulcers were especially difficult-to-heal, multifactorial ulcers with two or more etiologic factors. The treatment was more successful when the ulcers were smaller and had existed for a shorter period. One of the 10 patients had an amputation due to ulcers by diabetes, arterial, and pyoderma gangrenosum components. In the authors’ study group of ulcers of various other causes, the HDS did not heal the three pyoderma gangrenosum ulcers even if used together with cyclosporine. This result is in contrast to the study by de Imus, et al.,[30] who reported hastened healing and diminished pain when they treated a newly diagnosed case of ulcerative pyoderma gangrenosum with bioengineered skin and cyclosporine. In fact, HDS also did not heal the ulcers in the patient with scleroderma, but it seemed very effective in the three cases of necrobiosis lipoidica without diabetes mellitus. Comparing the treatment of hard-to-heal ulcers with other treatment methods is difficult, because no large, comparative studies have been carried out. LSE successfully has been used in the treatment of 21 patients with chronic leg ulcers of various origins, and the report also expands the diseases for which LSE may have some utility as an adjunctive therapy.[31] If one considers the treatment with HDS with surgical intervention, it should be noted that not all patients are suitable for surgical interventions, and the resources are limited. Traditional skin grafting also requires the creation of a second surgical site, the donor site. Donor sites are painful when healing, subject to infections, may leave scars or pigmentary alteration, and can only be used a limited number of times. Surgical intervention is very effective especially when the underlying etiologic cause is corrected at the same time.[32] These operations are technically complicated and not available or suitable for all diabetic patients. It was also estimated in a Finnish study that the excision and skin grafting required an average hospital stay of 11 days and a postoperative wound care of four and a half months. Ulcers recurred in 17 percent of these patients.[33] In chronic wounds, the amounts of various growth factors have been shown to be reduced when compared to acute wounds, and this may explain some of the poor healing rates associated with pressure or diabetic ulcers.[34,35] On the other hand, the treatment of chronic wounds with different growth factors so far has not been proven effective. The failure of topically applied growth factors may be due to inadequate dosage of the growth factor or the necessity of a “cocktail” of growth factors to effectively influence the wound stimulation. It is also possible that excessive administration of one factor may inhibit the effects of other essential factors.[36,37] Since it is known that HDS produces different kinds of growth factors and cytokines, some of the success achieved in the treatment of various chronic wounds can be explained with this fact. The authors’ study has several limitations. Pain was not measured, but it seemed to be reduced among patients with vasculitic or rheumatic ulcers as was described in a study by Öien, et al.,[29] who used pinch grafting technique in the treatment of rheumatic ulcers. The authors did not evaluate the associations between wound healing and patients’ weight, cigarette smoking habits, the type or duration of diabetes mellitus, and the influence of comorbidities. On the other hand, the primary goal of this study was to assess the treatment benefit, not these associations. After grafting with HDS, the patient is no longer dependent on the daily changing of the dressings by the nurse and gains greater social mobility. It is also a painless, noninvasive method and so does not require anaesthesia and can be performed as an outpatient procedure. Although no disease transmission or rejection reactions have occurred with bioengineered skin substitutes, they still are a concern, as is the relatively high cost of HDS. In general, conservative and time-honored methods of wound care should be attempted first. The results of the study suggest that HDS is a useful adjuvant therapy in chronic skin ulcers of various origins. Further evaluation is needed. The efficacy and tolerability of the product must be confirmed in randomized, prospective, clinical trials where a single type of ulcer is investigated with also more emphasis on the cost effectiveness, pain measurement, and the quality of life. More information on patients at high risk of delayed healing for inclusion in future trials is also needed.