Intimal thickness associated with endothelial dysfunction in human vein grafts

Abstract

Background: Intimal hyperplasia is a complex process thought to be initiated by injury and is the leading cause of vein graft failure. In the present investigation, we hypothesized that the basal intimal thickness in the human saphenous vein is a predictor of endothelial dysfunction and, potentially, intimal hyperplasia.

Methods: Human saphenous veins were obtained during coronary artery bypass surgery. The segments were contracted with phenylephrine and relaxed with carbachol to determine the endothelial-dependent relaxation. The vein segments were fixed in 10% buffered formalin and grown for 14 d in high-serum culture and then fixed in formalin. The fixed tissues were stained with Verhoeff-Van Gieson, and the average intimal and medial thicknesses were calculated using light microscopy and a computerized image analysis system.

Results: The human saphenous veins displayed varying amounts of basal intimal thickness (range 18.80-241.3 μm). The endothelial-dependent relaxation of the veins was highly variable, with values ranging from 0% to 27.59%. Human saphenous veins with a basal intimal thickness greater than 120 μm had significantly less endothelial-dependent relaxation (8.90% ± 6.32%) than those with a basal intimal thickness less than 120 μm (21.97% ± 10.64%). Endothelial dysfunction correlated with a basal intimal thickness greater than 120 μm (P = 0.02). The basal intimal thickness also correlated with increased intimal thickness after 14 d in organ culture (P = 0.0001).

Conclusions: A basal intimal thickness greater than 120 μm is a predictor of endothelial dysfunction. Also, because a greater basal intimal thickness correlated with an increased intimal thickness after organ culture, the basal intimal thickness might predict vein graft failure owing to intimal hyperplasia.

FIGURE 1

Basal intimal thickening in human saphenous veins and left internal mammary artery. Human saphenous vein and left internal mammary artery rings were fixed in formalin, sectioned and stained with Verhoeff-Van Gieson stain and examined by microscopy. Representative images of two saphenous vein segments with thin (A, 5× image and B, 40× image) and thick (C, 5× image and D, 40× image) intima and a left internal mammary artery (E, 5X image and F, 40X image) stained with Verhoeff-Van Gieson. In the image: L=lumen, IEL=internal elastic lamina (arrow), M=media, white line=thickness of intima. Scale bar=100 µm for 5× images and 50 µm for 40X images.

FIGURE 2

Variability of intimal thickening in Human saphenous veins and left internal mammary artery: Human saphenous vein (HSV, n=41) and left internal mammary artery (LIMA, n=4) rings were fixed in formalin, sectioned and stained with Verhoeff-Van Gieson stain and examined by microscopy. Scatter plot demonstrating the variability of basal intimal thickness measured as an average from two vein segments from each patient as described in the methods section.

FIGURE 3

Variability of endothelial-dependent relaxation in human saphenous veins. Human saphenous vein and left internal mammary artery segments were collected immediately after harvest and subjected to physiologic measurement in a muscle bath. Endothelial-dependent relaxation was measured by contracting with 10⁻⁶ M phenylephrine (PE) and relaxing with 5×10⁻⁷ M carbachol (Cch). A) Representative tracing of pre-contracted HSV relaxed with carbachol to 24.22%. B) Endothelial-dependent relaxation variability for saphenous veins (HSV) and left internal mammary artery (LIMA).

FIGURE 4

Endothelial-dependent relaxation of human saphenous vein correlates negatively with increase in basal intimal thickness. Human saphenous vein segments obtained from coronary artery bypass surgery were subjected to physiologic measurement of endothelial-dependent relaxation to carbachol in a muscle bath. Histological examination using the Verhoeff-Van Gieson stain was used for the visualization of the internal elastic lamina and the basal intimal thickening was measured. A) A linear regression of the percent endothelial-dependent relaxation as a function of basal intimal thickness was run yielding R²=0.2634, (P=0.02, n=20). B) Vein segments with basal intimal thickening greater than 120 µm had impaired endothelial dependent relaxation (P=0.0119, n = 20).

FIGURE 5

Intimal thickening in the basal state predicts degree of thickening in organ culture. Human saphenous vein segments were obtained from coronary artery bypass surgery and were cultured in RPMI medium with 30% serum for 14 days (post culture). Rings were fixed, sectioned, stained with Verhoeff-Van Gieson and analyzed by microscopy and intimal thickness was measured and compared to the intimal thickness before culture. A) Linear Regression of the change in intimal thickness from basal to post 14 day culture R² = 0.7794 (P<0.0001, n=21) shows a positive correlation where a higher basal intimal thickening leads to a higher intimal thickening in culture. B) Linear Regression of the change in intimal to medial ratio from basal to post 14 day culture R²=0.7587 (P<0.0001, n=21) also showing a positive correlation between basal intimal thickening and intimal thickening in culture.