Effect of mechanical fatigue on commercial bioprosthetic TAVR valve mechanical and microstructural properties

Abstract

Valvular structural deterioration is of particular concern for transcatheter aortic valve replacements due to their suspected shorter longevity and increasing use in younger patient populations. In this work we investigated the mechanical and microstructural changes in commercial TAVR valves composed of both glutaraldehyde fixed bovine and porcine pericardium (GLBP and GLPP) following accelerated wear testing (AWT) as outlined in ISO 5840 standards. This provided greater physiological relevance to the loading compared to previous studies and by utilizing digital image correlation we were able to obtain strain contours for each leaflet pre and post fatigue and identify sites of fatigue damage. The areas of greatest change in mechanical strain for each leaflet were then further probed using biaxial tensile testing, confocal microscopy, and electron microscopy. It was observed that overall strain decreased in the GLPP valves following AWT of 200 million cycles while the GLBP valve showed an increase in overall strain. Biaxial tensile testing showed a statistically significant reduction in stress for GLPP while no significant changes were seen for GLBP. Both confocal and electron microscopy showed a disruption to the gross collagen organization and fibrillar structure, including fragmentation, for GLPP but only the former for GLBP. However, further test data is required to confirm these findings and to provide a better understanding of this fatigue pathway is required such that it can be incorporated into both valve design and selection processes to improve overall longevity for both GLPP and GLBP devices.

Keywords: Biaxial tensile testing; Digital image correlation; Fatigue; Microstructure; TAVR valve.

Fig. 1.

Experimental arrangement for the DIC evaluation of valve strain under physiological pressures. (a) Test chamber aligned vertically with cameras whose position relative to the chamber can be altered by moving along the rails. (b) Valve in the test chamber; the valve is mounted in a silicone rubber ring which is held in rubber tubing which itself is held in a bearing which is mounted to the rear plate of the chamber with seals preventing leakage. The bearing allows the valve to be rotated without compromising the seal. (c) Alternate view of the full arrangement showing the rear of the test chamber; the tubing comes out of the rear of the chamber where it is plugged. The plug has an outlet for withdrawing fluid and creating the pressure difference across the valve and an outlet for measuring pressure. (d) Topdown view showing the ring backstop in the tubing which prevents valve migration when under pressure.

Fig. 2.

DIC strain fields pre (a) and post (b) AWT for the M26ER valve. A histogram of strain for the whole valve is adjacent to the scalebar.

Fig. 3.

DIC strain fields pre (a) and post (b) AWT for the M29EP + valve. A histogram of strain for the whole valve is adjacent to the scalebar.

Fig. 4.

DIC strain fields pre (a) and post (b) AWT for the ES23 valve. Mean strain values for each leaflet are placed on the center of each strain contour map. High values of strain or areas of interest are labeled with values that are away from the contour map to distinguish them from the mean values. A histogram of strain for the whole valve is adjacent to the scalebar.

Fig. 5.

Biaxial tension test data for M26ER (a), M29EP+ (b), ES23 (c), and M26ER and M29EP + combined (d) for control and fatigued samples (in areas of high strain change pre and post AWT). Median data is shown with error bars representing the 1st and 3rd quartiles

Fig. 6.

Representative confocal microscopy images of each respective sample control and post fatigue (in area of previously identified high strain change). Arrows have been added to aid in the identification of collagen fibers and their orientations. Images are 370 μm × 370 μm.

Fig. 7.

Representative TEM images of each respective sample control and post fatigue (in area of previously identified high strain change). Regions were selected to have fibrils plane and, for the fatigued samples, show the damage present to the fibrils. Each image is 13.3 μm × 13.3 μm.

Fig. 8.

Boxplot of segmented fibril lengths using CT-Fire software. Red star indicates a significant decrease between the left and right variables while a blue star indicates a significant increase (p < 0.05).

Fig. 9.

Contour plot of fiber angle versus slice number for the GLPP valves in the XY and XZ planes.

Fig. 10.

Contour plot of fiber angle versus slice number for the GLBP valve in the XY and XZ planes.