Dual Contrast‐Enhanced Microcomputed Tomography Uncovers Leaflet‐Dependent Patterns of Macrocalcification, Fibrous Extracellular Matrix Remodeling, and Adipocyte Degeneration in Calcific Aortic Stenosis

Figures

• 

  Figure 1. DCE‐μCT of calcium nodules and the trilayer structure in diseased and control leaflets. Representative DCE‐μCT images of (A) a diseased leaflet and (B) a control leaflet showing 3 layers of the valve—fibrosa, spongiosa, and ventricularis—and calcium nodules only in the diseased leaflet. C, Representative DCE‐μCT images of 3 types of calcium nodules seen in the diseased leaflets, with calcium nodules (normal arrows) and osmium‐labeled inclusions (ray‐end arrows) indicated. D, A bar graph plotting the prevalence of different types of calcium nodules. E, A bar graph (mean±SD) comparing the calcification volume fractions of male (white bars; n=30) and female (gray bars; n=15) diseased leaflets. P value from Student’s t test (E; *, significance with P<0.05). CN indicates calcium nodule; DCE‐μCT, dual contrast‐enhanced microcomputed tomography; F, fibrosa; S, spongiosa; and V, ventricularis.

• 

  Figure 2. Comparison of DCE‐μCT and histology for assessing trilayer thicknesses of valve leaflets. Representative DCE‐μCT (left of each panel) and histology (right of each panel) images of (A) a diseased and (B) a control leaflet at the same locations near midline showing similar fibrosa and ventricularis thicknesses. Scatter plots of DCE‐μCT‐derived vs histology‐derived thicknesses for (C) fibrosa, (D) ventricularis, (E) and spongiosa (n=18 measurements per valve layer), with best‐fit line and correlation coefficient (r) shown for each plot. DCE‐μCT indicates dual contrast‐enhanced microcomputed tomography.

• 

  Figure 3. DCE‐μCT of bright inclusions in control leaflets. A, Representative images of bright inclusions (arrows) in control valve leaflets from 19‐year‐old, 44‐year‐old, and 73‐year‐old subjects. B, Representative μCT images of 2 halves of a control leaflet at the same midline cut location. The left half was double stained with UA and OsO4 and the right half stained with UA only. C, Representative μCT images of 2 halves of a control leaflet at the same midline cut location. The left half was double stained with UA and OsO4 and the right half stained with OsO4 only. D, Bar graphs (mean±SD) comparing the CNRs of fibrosa on μCT performed after different staining protocols indicated were executed on a combined set of diseased (n=6) and control (n=6) leaflets. E, Bar graphs (mean±SD) comparing the CNRs of ventricularis on μCT performed after different staining protocols indicated were executed on the same set of diseased and control leaflets. P values from paired t test (D, E; *, significance with P<0.05). CNR indicates contrast‐to‐noise ratio; DCE‐μCT, dual contrast‐enhanced microcomputed tomography; DS, double staining; OsO4, osmium tetroxide; and UA, uranyl acetate.

• 

  Figure 4. SEM, EDS, TEM, and histology of bright inclusions in control leaflets. Representative SEM images at (A) low and (B) high magnification showing clusters of bright inclusions in the spongiosa, with (C) some inclusions having small droplet(s) in the periphery (arrow). D, Representative EDS of an inclusion showing characteristic osmium peaks, in addition to background carbon and oxygen peaks. E, TEM of a bright inclusion showing a nucleus (arrow) in the periphery of a large central lipid droplet (*). On histology, these inclusions appear as white voids (arrows) on (F) hematoxylin and eosin and (G) Movat staining and (H) express perilipin‐1 (green fluorescence). C indicates carbon; EDS, energy dispersive spectroscopy; O, oxygen; Os, osmium; SEM, scanning electron microscopy; and TEM, transmission electron microscopy.

• 

  Figure 5. SEM, TEM, and histology of adipocytes in diseased leaflets. Representative (A) low and (B) high magnification SEM images of a diseased leaflet showing calcium nodules and scant adipocytes (arrows) nearby. In selected areas, MFNs can be seen as (C) flat entities with fluffy contour on SEM. D, On histology, MFNs often appear next to adipocytes in the spongiosa (black box) as (E) purplish thick‐contoured species on Movat staining (arrows), (F) devoid of perilipin‐1 staining (green fluorescence; adipocyte marker). Near calcium nodules, clusters of osmiophilic materials can sometimes be observed with (G) irregular sizes, shapes, and contours, which on TEM appear as (H) spots of acellular and homogeneous background, representing necrotic remnants. CN indicates calcium nodule; MFN, membranous fat necrosis; SEM, scanning electron microscopy; and TEM, transmission electron microscopy.

• 

  Figure 6. DCE‐μCT‐based feature analysis of OLIs in valve leaflets with SEM correlations. Tukey box plots comparing the (A) bulk volume fraction, (B) bulk normalized surface area, and (C) voxel intensity of OLIs (all assessed by DCE‐μCT) in control (n=24; white bars) vs diseased leaflets (n=24; gray bars). D, Correspondingly, DCE‐μCT (left) and SEM (middle) images of the same location in a control leaflet, correlating OLIs (arrows) with adipocytes (ray‐end arrows), confirmed at high magnification (right). E, Corresponding DCE‐μCT (left) and SEM (middle) images of the same location in a diseased leaflet, correlating OLIs (arrows) more often with MFNs (ray‐end arrows), best shown at high magnification (right). Tukey box plots comparing the SEM‐derived (F) adipocyte and (G) MFN fractions in control (n=18; white bars) and diseased (n=18; gray bars) leaflets. P values from Wilcoxon rank‐sum test (A, B, C, F, G; *, significance with P<0.05). DCE‐μCT indicates dual contrast‐enhanced microcomputed tomography; MFN, membranous fat necrosis; OLI, osmium‐labeled inclusion; and SEM, scanning electron microscopy.

• 

  Figure 7. Correlating spatial distributions of OLIs with calcification in control and diseased leaflets. A, Convention used for describing different regions of a valve leaflet, with each leaflet divided into the coaptation region and equally the upper and lower noncoaptation regions for convenient segmentation in silico. Note that, in the center of the leaflet, these regions correspond to the coaptation, belly, and hinge regions. B, Representative scout DCE‐μCT image of a control leaflet with OLIs (green) overlaid and (C) a diseased leaflet with calcification (red) and OLIs (green) overlaid. D, Tukey box plots of OLI volume fraction associated with different regions in control leaflets (n=24). E, Tukey box plots of OLI volume fraction and (F) calcification volume fraction associated with different regions in diseased leaflets (n=24). P values from Wilcoxon signed‐rank test (D, E, F; *, significance with P<0.0167). Coapt indicates coaptation; DCE‐μCT, dual contrast‐enhanced microcomputed tomography; LNC, lower noncoaptation; OLI, osmium‐labeled inclusion; and UNC, upper noncoaptation.