In diabetic patients, wound healing is impaired. We studied the pathogenesis behind this clinical observation by characterizing the pattern of deposition of extracellular matrix (ECM) molecules and the cellular infiltrate in chronic (>8 wk) diabetic wounds, compared with chronic venous ulcers and an acute wound healing model. Punch biopsies were obtained from the chronic ulcer margins and control samples were collected from upper leg skin 5, 19, 28 d and 12 and 18 mo postwounding (p.w.). T cells, B cells, plasma cells, granulocytes and macrophages, and the ECM molecules fibronectin (FN), chondroitin sulfate (CS), and tenascin (TN) were visualized using immunohistochemical techniques. Expression of FN, CS, and TN was detected in dermal tissue early in normal wound healing (5-19 d p.w.). Abundant staining was seen 3 mo p.w., returning to prewounding levels after 12-18 mo p.w. In the dermis of chronic diabetic and venous ulcers with a duration of 12 mo or more, a prolonged presence of these ECM molecules was noted. Compared with normal wound healing: (i) the CD4/CD8 ratio in chronic wounds was significantly lower (p < 0.0027) due to a relatively lower number of CD4+ T cells; (ii) a significantly higher number of macrophages was present in the edge of both type of chronic ulcers (p < 0.001 versus day 29 p.w.); and (iii) more B cells and plasma cells were detected in both type of chronic wounds compared with any day in the acute wound healing model (p < 0.04 for CD20+ and p < 0.01 for CD79a+ cells). These data indicate that important differences exist in the cellular infiltrate and ECM expression patterns of acute, healing versus chronic wounds, which may be related to the nonhealing status of chronic wounds.

Our results illustrate that numbers of fibroblasts in the collagen matrix and their functional state is a critical factor for establishment of normal epidermal morphogenesis.

Patients with diabetes mellitus experience impaired wound healing often resulting in chronic foot ulcers. Hospital discharge data indicate that 6-20% of all diabetic individuals hospitalized (mostly with type 2 diabetes) have a lower extremity ulcer. Maintaining glucose levels at acceptable levels (below 10 mmol/l) is considered to be an important part of the clinical treatment, but the exact mechanism by which diabetes delays wound repair is not yet known. We studied this phenomenon by determining the potential of fibroblasts isolated from the ulcer sites of four patients with non-insulin-dependent diabetes mellitus to proliferate in vitro. Controls were fibroblasts isolated from normal skin of the upper leg of five healthy age-matched volunteers and of six non-insulin-dependent diabetes patients. Proliferative capacity was analysed by evaluation of plates after trypsinization and [3H]thymidine incorporation. Fibroblast morphology was studied by light and transmission electron microscopy. Diabetic ulcer fibroblasts, measured by [3H]thymidine incorporation, proliferated significantly more slowly than the nonlesional control fibroblasts (P < 0.00047) and age-matched control fibroblasts (P < 0.00003). After culturing the fibroblasts for a prolonged period in high-glucose (27.5 mM) and low-glucose (5.5 mM, i.e. physiological) medium, this difference in proliferation rate between diabetic ulcer fibroblasts and nonlesional diabetic fibroblasts remained (P < 0.0001 for high-glucose and P < 0.0009 for low-glucose on day 7). Fibroblast proliferation in all three groups was slightly lower in high-glucose than in low-glucose medium, although not significantly at any time-point. Light microscopy showed diabetic ulcer fibroblasts to be large and widely spread. Transmission electron microscopy of cultured diabetic ulcer fibroblasts and nonlesional diabetic skin fibroblasts revealed a large dilated endoplasmic reticulum, a lack of microtubular structures and multiple lamellar and vesicular bodies. These results show a diminished proliferative capacity and abnormal morphology of fibroblasts derived from diabetic ulcers of non-insulin-dependent diabetes patients.

Poly(ether ester) block-copolymer scaffolds of different pore size were implanted into the dorsal skinfold chamber of balb/c mice. Using intravital fluorescent microscopy, the temporal course of neovascularization into these scaffolds was quantitatively analyzed. Three scaffold groups (diameter, 5 mm; 220 -260 thickness, m; n ϭ 30) were implanted. Different pore sizes were evaluated: small (20 -75 m), medium (75-212 m) and large pores (250 -300 m). Measurements were performed on days 8, 12, 16, and 20 in the surrounding normal tissue, in the border zone, and in the center of the scaffold. Standard microcirculatory parameters were assessed (plasma leakage, vessel diameter, red blood cell velocity, and functional vessel density). The largepored scaffolds showed significantly higher functional vessel density in the border zone and in the center (days 8 and 12) compared with the scaffold with the small and mediumsized pores. These data correlated with a larger vessel diameter and a higher red blood cell velocity in the largepored scaffold group. Interestingly, during the evaluation period the microcirculatory parameters on the edge of the scaffolds returned to values similar to those found in the surrounding tissue. In the center of the scaffold, however, neovascularization was still active 20 days after implantation. Plasma leakage and vessel diameter were higher in the center of the scaffold. Red blood cell velocity and functional vessel density were 50% lower than in the surrounding tissue. In conclusion, the dorsal skinfold chamber model in mice allows long-term study of blood vessel growth and remodeling in porous biomedical materials. The rate of vessel ingrowth into poly(ether ester) block-copolymer scaffolds is influenced by pore size and was highest in the scaffold with the largest pores. The data generated with this model contribute to knowledge about the development of functional vessels and tissue ingrowth into biomaterials.