Dynamic Luminal Topography: A Potential Strategy to Prevent Vascular Graft Thrombosis

Abstract

Aim: Biologic interfaces play important roles in tissue function. The vascular lumen-blood interface represents a surface where dynamic interactions between the endothelium and circulating blood cells are critical in preventing thrombosis. The arterial lumen possesses a uniform wrinkled surface determined by the underlying internal elastic lamina. The function of this structure is not known, but computational analyses of artificial surfaces with dynamic topography, oscillating between smooth and wrinkled configurations, support the ability of this surface structure to shed adherent material (Genzer and Groenewold, 2006; Bixler and Bhushan, 2012; Li et al., 2014). We hypothesized that incorporating a luminal surface capable of cyclical wrinkling/flattening during the cardiac cycle into vascular graft technology may represent a novel mechanism of resisting platelet adhesion and thrombosis.

Methods and results: Bilayer silicone grafts possessing luminal corrugations that cyclically wrinkle and flatten during pulsatile flow were fabricated based on material strain mismatch. When placed into a pulsatile flow circuit with activated platelets, these grafts exhibited significantly reduced platelet deposition compared to grafts with smooth luminal surfaces. Constrained wrinkled grafts with static topography during pulsatile flow were more susceptible to platelet accumulation than dynamic wrinkled grafts and behaved similar to the smooth grafts under pulsatile flow. Wrinkled grafts under continuous flow conditions also exhibited marked increases in platelet accumulation.

Conclusion: These findings provide evidence that grafts with dynamic luminal topography resist platelet accumulation and support the application of this structure in vascular graft technology to improve the performance of prosthetic grafts. They also suggest that this corrugated structure in arteries may represent an inherent, self-cleaning mechanism in the vasculature.

Keywords: compliance; dynamic topography; ex vivo pulsatile flow model; platelets; prosthetic vascular graft.

FIGURE 1

Fabrication of dynamic silicone wrinkled graft. (A) Representative confocal image of a cross-section of pig carotid artery fixed under luminal pressure of 60 mmHg demonstrates the wrinkled configuration of the internal elastic lamina (green autofluorescence for elastin), resulting in a wrinkled luminal surface lined by endothelial cells. (B) Schematic representation of the fabrication of a bilayer silicone graft with a wrinkled inner lining under resting conditions. A thin film composed of a stiffer silicone is dip-coated on a 6.3 mm mandrel. An outer tube composed of a soft silicone is casted on a 3.2 mm mandrel. This soft tube is removed from the mandrel and stretched over the stiff film on the 6.3 mm mandrel. Removal of this bilayer graft from the larger mandrel results in recoil of the outer tube. The stiffer inner film buckles into a uniform wrinkled luminal surface. This bilayer graft can distend with pressure which results in flattening of the luminal surface. Under lower pressures, the graft relaxes and the luminal wrinkles reform. (C) Photograph of the bilayer graft showing luminal corrugations. (D) Representative photomicrograph of a cross section of the graft showing the wrinkled inner film. (E) Representative scanning electron microscopy of the luminal surface of wrinkled grafts showing the smooth characteristics of the silicone surface without any nanoscale topography.

FIGURE 2

Effect of luminal film on graft wrinkle size and graft compliance. Grafts were fabricated with a thin stiff luminal film that determined the size of the luminal wrinkles when the outer, more elastic tube recoiled. (A) Sample cross-sectional differential interference contrast (DIC) microscopy images of wrinkled grafts are shown and focus on the luminal film layer. The thickness of the luminal film determines the size of the wrinkles with thicker films creating larger wrinkles. (B) The thickness of the luminal film and the wavelength were measured from DIC images of eight different grafts and plotted. The graph shows that wavelength increases proportionately with film thickness. Error bars are the standard deviations in wrinkle wavelength and luminal film thickness from several measurements taken along each individual graft. (C) Grafts with wrinkled and smooth luminal topography with the same luminal film thickness were tested for compliance. Inflation characteristics of these grafts were examined under continuous flow. The outer tube without the luminal film was evaluated for comparison only; no other experiment was performed with the outer tube. The % change in strain is plotted versus luminal pressure. The grafts with wrinkled and or smooth luminal film have similar inflation characteristics (P = NS at all pressures), whereas the tube has higher distensibility (N = 4 grafts/condition; ^∗P < 0.001 between outer tube vs. smooth and wrinkled grafts; ^∗∗P = 0.049, outer tube vs. wrinkled graft and P = 0.1 outer tube vs. smooth graft; ^†P < 0.067 outer tube vs. wrinkled and smooth grafts; ‡ P = 0.04 outer vs. smooth grafts and P = 0.087 outer vs. wrinkled grafts).

FIGURE 3

Dynamic topography of wrinkled grafts and effect on platelet adherence. Grafts with wrinkled luminal surfaces were placed in a pulsatile flow circuit with peak pressures of 140 mmHg. Representative OCT images reveal progressive flattening of the luminal surface wrinkles as luminal pressure increased (repeated with four different grafts). The timeline at the bottom of each image shows a pulse frequency of 36 per minute and the reproducible distention of the graft in that location with each pulse (A). These experiments were performed using wrinkled grafts with ∼200 μm wavelength to allow better visualization on OCT to demonstrate dynamic topography. Smooth and wrinkled bilayer grafts were placed on a pulsatile flow pump with thrombin activated platelets. Peak pressures were set for 140 mmHg while the resting pressure was 60–80 mmHg to correlate with diastolic pressures in vivo. The pulse rate was set to 40 beats/minute. Grafts were removed at 1 h and stained with Wright stain to detect accumulated platelets. Platelet deposition per high powered field (HPF) was quantified using ImageJ software. Representative photomicrographs revealed that smooth grafts (B) exhibited significantly more adherent platelets (blue dots and arrows) as compared with wrinkled grafts (wrinkle wavelength ∼100 μm) (C). Summary of the results for the experiment are shown graphically (N = 4 grafts/treatment group, 10 HPFs were quantified per graft) (D). The comparison of platelet accumulation between smooth and wrinkled grafts was conducted using Wilcoxon signed Rank test.

FIGURE 4

Effect of luminal surface behavior on platelet deposition. (A) To test the role of dynamic topography in resisting platelet adherence, a portion of the dynamic graft was externally constrained to prevent graft distension during pulsatile flow to maintain a fixed wrinkled luminal surface under pulsatile flow. The adjacent dynamic portion of graft continued to undergo distension during pulsatile flow and cycled between wrinkled and smooth luminal configurations. (B) Grafts were placed on pulsatile pump conditions with thrombin activated platelets for 1 h, stained with Wright stain, and platelet deposition per high powered field was quantified using Image J software. Externally constrained wrinkled grafts (n = 4–5 grafts/condition) demonstrated increased platelet deposition compared with dynamic wrinkled grafts (P < 0.035). Similarly, constrained smooth grafts also had increased platelet deposition compared to dynamic smooth grafts (P < 0.021). Dynamic wrinkled grafts were more resistant to platelet deposition than smooth dynamic grafts (P < 0.001). Constrained wrinkled grafts were also more resistant to platelet accumulation than constrained smooth grafts (P < 0.001). Comparisons were performed with one-way ANOVA with Dunn’s test. (C) Wrinkled grafts (∼100 μm wavelength, n = 3 grafts) were either placed on pulsatile (dynamic topography) or continuous (fixed topography) flow conditions for 1 h with activated platelets. Grafts undergoing pulsatile flow had significantly lower platelet adherence (P < 0.001) than those exposed to continuous flow. The comparison was performed with the Wilcoxon signed-rank test for pair wise comparisons. (D) Grafts with different wavelength wrinkles were placed under pulsatile flow with activated platelets. Comparisons were performed with one-way ANOVA with Dunn’s test for multiple comparison. In comparison with grafts with 300 μm wrinkles, those with smaller wrinkle sizes were significantly more resistant to platelet aggregation (N = 3–4 grafts/wavelength; ^∗P < 0.001). Grafts with 100 μm wrinkles showed a trend toward a significant reduction in platelet adherence compared to grafts with 150 μm wrinkles (^†P = 0.073) and significantly reduced platelet adherence versus grafts with 200 μm wrinkles (^‡P = 0.005).