Architectural trends in the human normal and bicuspid aortic valve leaflet and its relevance to valve disease

Abstract

The bicuspid aortic valve (AV) is the most common cardiac congenital anomaly and has been found to be a significant risk factor for developing calcific AV disease. However, the mechanisms of disease development remain unclear. In this study we quantified the structure of human normal and bicuspid leaflets in the early disease stage. From these individual leaflet maps average fiber structure maps were generated using a novel spline based technique. Interestingly, we found statistically different and consistent regional structures between the normal and bicuspid valves. The regularity in the observed microstructure was a surprising finding, especially for the pathological BAV leaflets and is an essential cornerstone of any predictive mathematical models of valve disease. In contrast, we determined that isolated valve interstitial cells from BAV leaflets show the same in vitro calcification pathways as those from the normal AV leaflets. This result suggests the VICs are not intrinsically different when isolated, and that external features, such as abnormal microstructure and altered flow may be the primary contributors in the accelerated calcification experienced by BAV patients.

FIGURE 1

From left: in vivo non-calcified TAV, explanted calcified TAV, explanted non-calcified and calcified BAV leaflets. Calcification is usually coupled with strong inflammation and thickening.

FIGURE 2

(a) Collagen architecture visible in TAV and BAV leaflets under polarized light and (b) quantified fiber architecture for representative TAV and BAV leaflets using SALS setup.

FIGURE 3

Method of spline flitting: Parameter space is specialized for the aortic valve by degenerating two of its edges. This gives a physical relation between the two configurations of the valve (2D and 3D) and a mapping in between them.

FIGURE 4

Steps in calculating the average microstructure. SALS experiment for each leaflet gives an individual microstructure map. These maps are fitted with spline curves and converted to surfaces. Then using the spline mapping, the SALS data is mapped onto a common template.

FIGURE 5

(a) The projection or closest point is orthogonal to the curve/surface and is calculated by solving the normality condition using Newton’s method. (b) In case of surface, some boundary points may not satisfy this condition fully and have to be constrained on the boundary.

FIGURE 6

Spline fitting to a 3D point cloud includes an extra step after the conversion of spline curve into a surface: the interior control points are fitted to minimize the error and obtain the final spline representation (color represents the projection error).

FIGURE 7

(a) Analysis of aortic valve (AV) microstructure, ECM organization in representative tissues from non calcified and calcified tricuspid (TAV) and BAV—H&E and Movat staining in the AV tissues; ECM components are distinguished as proteoglycan (bluish green), collagen (yellow), and elastin (dark violet); VIC nuclei are stained dark red; F, S and V represent fibrosa, spongiosa and ventricularis respectively. (b, c) Characterization of VICs isolated from AV tissues—western blot and bar graph showing SMA and vimentin expression in AV VICs, GAPDH serves as loading control. (d) In vitro calcification assay—images are representative of calcification assay performed on VICs isolated from non-calcified TAV and BAV using 3 mM Pi. Calcium was quantified by colorimetric assay and represented as calcium mg/mL.

FIGURE 8

(a) Representative layer arrangement from histology shows organized and aligned layers in the TAV but disorganized ones in BAV especially around the raphe. (b) A packet of elastin is seen in the BAV raphe region.

FIGURE 9

(a) Basic geometric parameters—distance between commissure points d, length of the free edge L_F, length of the basal attachment L_B and area S, were calculated for all the explants and divided into two groups, TAV and BAV, for statistical analysis. There were statistically significant differences in all four parameters, (b) average geometry of TAV and BAV using these parameters (*p <0.01 and **p <0.005).

FIGURE 10

(a) Explants from human sub-population TAV and BAV were used to determine the microstructure and then averaged to obtain the average microstructure maps. (b) Statistical analysis on mean fiber directions (left) and OI (right) shows important differences in the belly region for OI but the mean fiber directions are similar in two cases. (c) Summary statistics based on regions shows consistency in the structure of the leaflets from both groups (*p <0.01 and **p <0.005).

FIGURE 11

Using the 3D spline mapping, average microstructure was mapped onto in vivo geometries for TAV and BAV respectively (obtained from hand-digitized ultrasound imaging of human aortic valve at mid-systole point of cardiac cycle).