Increased but ineffectual angiogenic drive in nonhealing venous leg ulcers

Objective: Our previous work demonstrated that angiogenesis is inhibited in nonhealing venous ulcers. The object of this study was to determine whether local expression of vascular endothelial growth factor (VEGF) and other major regulators of vessel growth are related to healing of venous ulcers. Subjects and methods: The study included 35 patients with venous ulcers (CEAP 6) and 9 patients whose ulcers had healed (CEAP 5). Control subjects were 18 patients undergoing routine operations (8 with closed suction drains, 10 standard skin biopsies). Healing ulcers were defined as having healed in less than a year from entry to the study; nonhealing ulcers failed to heal in this period. A 1-cm square biopsy specimen was taken from the edge of the ulcer or from a site of lipodermatosclerosis around a healed ulcer. Wound fluids were aspirated from beneath transparent occlusive dressings. Concentrations of VEGF(165) and VEGF-R1 were measured in tissue homogenates with enzyme-linked immunosorbent assay, and results are expressed as mean +/- SEM per milligram of soluble protein (SP). Expression of mRNA transcripts for the VEGF splice variants VEGF(121), VEGF(189), and VEGF(165); the receptors VEGF-R1 and VEGF-R2; the angiopoietins Ang-1 and Ang-2; and their receptor, Tie-2, were measured in biopsy samples with multiplex polymerase chain reaction. Expression of each transcript was normalized to that of the housekeeping gene, GAPDH. Results were analyzed with analysis of variance, t test, and chi(2) test. Results: There was no difference in VEGF(165) protein concentration between biopsy specimens from healing ulcers (2.12 +/- 0.34 ng/mg SP; n = 18) and nonhealing ulcers (2.36 +/- 0.39 ng/mg SP; n = 12), but concentration was higher in all ulcer samples compared with healthy skin (0.57 +/- 0.20 ng/mg SP; n = 10; P < .01) and healed ulcers (0.33 +/- 0.06 ng/mg SP; n = 9; P < .01). Concentration of VEGF(165) protein in wound fluid was significantly higher in nonhealing venous ulcers (67.17 +/- 13.87 ng/mg SP; n = 13) compared with healing venous ulcers (32.19 +/- 7.90 ng/mg SP; n = 19; P <.05) or acute wounds (12.26 +/- 4.50; n = 8; P <.01). Concentration of VEGF-R1 was similar in wound fluid obtained from healing ulcers (7.18 +/- 1.34 ng/mg SP; n = 13) and nonhealing ulcers (7.02 +/- 1.21 ng/mg SP; n = 19), and acute wounds (7.12 +/- 2.35 ng/mg SP; n = 8). There was a weak but significant correlation between VEGF(165) protein concentration in the ulcer biopsy specimen and wound fluid from the same ulcer (R-2 = 0.2; P =.019; n = 27). Expression of mRNA for VEGF receptors and Tie-2 was poor. VEGF121 was expressed in all samples, and VEGF165 in 43 of 48 samples. mRNA expression of VEGF(189) (P=.001), Ang-1 (P=.002), and Ang-2 (P=.026) was found in more samples from unhealed ulcers than from other sites. Healed ulcers had reduced mRNA expression of VEGF(165) (0.181 +/- 0.003) than did healing ulcers (0.307 +/- 0.016; P =.007) or nonhealing ulcers (0.375 +/- 0.033; P =.001). Relative expression of VEGF(165) to Ang-2 was much lower in healed ulcers (0.4236 +/- 0.060) than in healing ulcers (1.382 +/- 0.235; P =.010) and nonhealing ulcers (1.887 +/- 0.280; P =.003). Conclusion: In nonhealing venous ulcers there is a consistently high level of expression of VEGF, at both the gene transcript and protein level. As our previous data demonstrated that angiogenesis is depressed in these poorly healing ulcers, an increase in VEGF production may indicate an increased but ineffectual angiogenic drive. It is also possible that undiscovered inhibitors are released in the ulcer environment.