Improved survival of ischemic cutaneous and musculocutaneous flaps after vascular endothelial growth factor gene transfer using adeno-associated virus vectors

Abstract

A major challenge in reconstructive surgery is flap ischemia, which might benefit from induction of therapeutic angiogenesis. Here we demonstrate the effect of an adeno-associated virus (AAV) vector delivering vascular endothelial growth factor (VEGF)165 in two widely recognized in vivo flap models. For the epigastric flap model, animals were injected subcutaneously with 1.5 x 10(11) particles of AAV-VEGF at day 0, 7, or 14 before flap dissection. In the transverse rectus abdominis musculocutaneous flap model, AAV-VEGF was injected intramuscularly. The delivery of AAV-VEGF significantly improved flap survival in both models, reducing necrosis in all treatment groups compared to controls. The most notable results were obtained by administering the vector 14 days before flap dissection. In the transverse rectus abdominis musculocutaneous flap model, AAV-VEGF reduced the necrotic area by >50% at 1 week after surgery, with a highly significant improvement in the healing process throughout the following 2 weeks. The therapeutic effect of AAV-VEGF on flap survival was confirmed by histological evidence of neoangiogenesis in the formation of large numbers of CD31-positive capillaries and alpha-smooth muscle actin-positive arteriolae, particularly evident at the border between viable and necrotic tissue. These results underscore the efficacy of VEGF-induced neovascularization for the prevention of tissue ischemia and the improvement of flap survival in reconstructive surgery.

Figure 1

Schematic representation of the skin flaps and their vascular components, with the indication of the vector injection sites. A: The surgical models of skin flap used in this study are based on a rectangular skin paddle measuring 5 × 8 cm, drawn on the abdomen of the animals. The predictable vascular system of the flap is composed of the lateral thoracic arteries, the inferior epigastric arteries, and the musculocutaneous perforator arteries arising from the rectus abdominis (usually four vessels on each side). B and C: The pictures schematically show the vascular component providing the blood supply to each flap and the injection sites. In particular, the skin flap (B) was raised from the fascia on the inferior epigastric artery, and the vector injected at 10 equally spaced subcutaneous sites along the midline (inset in the top left part). The TRAM flap (C) was raised on a plane between the panniculus carnosus and the abdominal fascia, with the rectus abdominis as the only source of blood supply through the perforator arteries; in this model, the vector was administered by intramuscular injection in the region where each perforator artery arises from the rectus sheet (inset in the top left part).

Figure 2

Effect of VEGF on skin flap survival at different times of administration. At postoperative day 7, the regions of survival and necrosis were clearly defined in all of the flaps: the surviving skin appeared pink and tender, whereas the distal necrotic portion was black and rigid. A: The pictures show two selected epigastric flaps, treated with AAV-LacZ (left) and AAV-VEGF (right), 14 days before flap elevation. The viability of the skin appears clearly improved after AAV-VEGF administration. B: The effect of AAV-VEGF on epigastric flap survival was assessed by injecting the vector at different time points before surgery (0, 7, and 14 days). The histograms represent the mean values and SD of the necrotic area relative to the total flap area, as measured by digital planimetry, for each experimental group. C: The pictures show two selected TRAM flaps, treated with AAV-LacZ (left) and AAV-VEGF (right), 14 days before flap elevation. Also in this model, the treatment with AAV-VEGF resulted in a significant improvement in tissue viability. D: The effect of AAV-VEGF on TRAM flap survival was assessed by injecting the vector at different time points before surgery (0, 7, and 14 days). The histograms represent the mean values and SD of the necrotic area relative to the total flap area, as measured by digital planimetry, for each experimental group.

Figure 3

Pretreatment with AAV-VEGF significantly improves TRAM flap survival and healing. A: The histogram shows the mean and SD of percent flap necrosis measured in 40 rats (20 per group) that were treated with the TRAM flap and injected with either AAV-LacZ or AAV-VEGF at day 14 before flap elevation. Measurements were performed at day 7 after surgery. B: The healing process was monitored throughout time after surgery up to 22 days, by measuring the extent of flap survival in animals treated with AAV-LacZ or AAV-VEGF (n = 8 per group). The means and SD of survived flap areas are shown at each time point (expressed as a percentage of the total flap area).

Figure 4

Histological sampling and assessment of TRAM flap viability. A: Each flap presented three distinguishable zones, according to their distance from the vascular pedicle: a survived zone (a), an intermediate zone (b), and a necrotic zone (c). To histologically examine flap viability, we harvested the skin and the muscular sheet from the injection site (sample A), as well as a more distal cutaneous sample (sample B, at ∼3 cm from sample A) from the intermediate zone. B: Shown are representative sections of samples A and B from AAV-VEGF-treated (right) and control (left) animals. At the injection site (sample A), a massive cellular infiltration appeared as a consequence of AAV-VEGF treatment (top). More notably, sample B of VEGF-treated flaps showed an intact and viable epithelial layer with conserved tissue architecture, whereas, in control flaps, the epithelium was thin and discontinuous, with massive inflammation and adipose substitution. Myonecrosis was detected only in LacZ-treated flaps, as indicated by the disappearance of the panniculus carnosus (shown by asterisks in the VEGF sample). Note the presence of circulating inflammatory cells in the arterial lumen in the insets, more abundant in the LacZ-treated as compared to the VEGF-treated samples.

Figure 5

VEGF induces endothelial cell proliferation and angiogenesis. A: The presence of endothelial cells in control (left) and VEGF-treated (right) flaps was detected by immunohistochemistry using an anti-CD31 antibody. AAV-VEGF induced the proliferation of endothelial cells at the injection site (sample A), in which several CD31-positive cells infiltrated the interstitial spaces between the fibers of the rectus abdominis muscle (inset on the right), as well as in the more distal sample B. This endothelial cell proliferation was paralleled by the formation of a great number of new capillaries, most evident at the level of the panniculus carnosus (p.c.), as shown in the bottom panels at a higher magnification. B: Quantification of capillaries in treated flaps. Counts were obtained from samples B of both AAV-LacZ- and AAV-VEGF-treated flaps. Shown are the means and SD of the counts, expressed as number of capillaries per ×40 field, assuming statistical significance at P < 0.05 using a two-tailed t-test.

Figure 6

VEGF induces the formation of α-SMA-positive arteriolae. A: The property of VEGF to sustain the formation of arterial vessels was assessed by immunohistochemistry using an antibody against the α-actin isoform specific for the smooth muscle cells (α-SMA). A modest effect was detected in sample A, with a greater reactivity in the AAV-VEGF-injected muscle (inset on the right), whereas a remarkable increase in the number of arteriolae was observed after VEGF treatment in the more distal sample B (bottom), approximately corresponding to the border between viable and necrotic skin. B: Quantification of α-SMA-positive vessels in samples B. Presentation of the data and statistics are as in Figure 5B.