Association of Coronary Microvascular Rarefaction and Myocardial Fibrosis With Coronary Artery Disease

Abstract

Background: To evaluate, in a cohort study, whether coronary microvasculature and myocardial structure differ between people with and without coronary artery disease (CAD).

Methods and results: We performed histological analysis of left ventricle free wall obtained at autopsy from 25 men and 23 women with ≥1 coronary artery with ≥75% area stenosis, and 25 men and 25 women without (no or minimal) CAD, matched for sex and age, who died suddenly from noncardiac causes. Decedents with myocardial infarction or other cardiac abnormality were excluded. Decedents with and without CAD had similar height and weight. Heart weight of decedents with CAD was higher than that of decedents without CAD (mean, 391 versus 364 g; mean difference, 27 g [95% CI, 0.3-54.0], P=0.048). Decedents with CAD had lower arteriole density (mean, 1.4 per mm² versus 1.8 per mm²; mean difference, -0.4 per mm² [95% CI, -0.6 to -0.2], P=0.0001), lower capillary length density (mean, 3164 versus 3701 mm/mm³; mean difference, -537 [95% CI, -787 to -286], P<0.0001), and higher total myocardial fibrosis (mean, 7.5% versus 5.7%; mean difference, 1.7% [95% CI, 1.0-2.5], P<0.0001), than decedents without CAD.

Conclusions: CAD was associated with coronary microvascular rarefaction and increased myocardial fibrosis. The association of CAD with coronary microvascular rarefaction and increased myocardial fibrosis may contribute to the increased risks of death, myocardial infarction and heart failure that accompany CAD, and may attenuate the impact of percutaneous coronary intervention on cardiovascular risk in people with stable angina.

Keywords: coronary artery disease; coronary microvasculature; myocardial fibrosis.

Figure 1. CONSORT diagram of strategy for search of National Coronial Information System to identify decedents with sudden noncardiac death and with or without coronary artery disease.

None of the decedents with CAD had complete occlusion of a coronary artery or intravascular thrombosis; one decedent with CAD had a small intra‐plaque hemorrhage. CAD indicates coronary artery disease; CONSORT, Consolidated Standards of Reporting Trials; MI, myocardial infarction; and PCI, percutaneous coronary intervention.

Figure 2. Picrosirius red‐stained section of left ventricle.

Digitized images of picrosirius red‐stained left ventricle section illustrating section area and length of section (A), and inner, middle, and outer regions of section (B). Left ventricle wall thickness was calculated by dividing the section area (262 mm²) by the section length (21.02 mm) to give left ventricle wall thickness (12.47 mm).

Figure 3. Reticulin‐stained section of left ventricle.

Digitized image of reticulin‐stained left ventricle section illustrating the outlines of cardiomyocytes, and measurement of the shortest diameter of cardiomyocyte profiles with a nucleus.

Figure 4. Picrosirius red‐stained section of left ventricle showing fibrosis.

Digitized image of picrosirius red‐stained left ventricle section illustrating fibrosis, stained red, and cardiomyocytes, stained orange. Total fibrosis of the inner, middle, and outer regions was quantified by digital analysis of picrosirius red‐stained sections using Aperio Imagescope software's Positive Pixel Count algorithm calibrated for hue and color saturation thresholds. Total fibrosis was calculated as the sum of the total pixel count for fibrosis, expressed as a percentage of the total pixel count for the tissue area for each region of each section.

Figure 5. Picrosirius red‐stained section of left ventricle showing arterioles and perivascular fibrosis.

Digitized image of picrosirius red‐stained left ventricle section illustrating fibrosis stained red and cardiomyocytes stained orange. Shown is an arteriole in approximate cross‐section, with internal wall (lumen area), external wall (vessel circumference and area), and perivascular fibrosis area outlined. Also shown are two arterioles and the measurement of their minimal diameters. Arteriole vessel area, perivascular fibrosis area, vessel circumference, and arteriole minimal diameter were quantified by digital analysis using Aperio Imagescope software. Perivascular fibrosis was calculated as the ratio of the area of perivascular fibrosis (total area of vessel and fibrosis minus total vessel area) and the total vessel area (area of vessel wall plus lumen). Arteriole wall area/circumference ratio was calculated as the ratio of wall area (total vessel area minus lumen area) and vessel circumference.

Figure 6. CD31 immuno‐stained section of left ventricle showing capillaries.

Digitized image of CD31 immuno‐stained left ventricle section illustrating capillaries, stained brown. Shown is a counting frame. Counting frames were superimposed over areas with minimally distorted tissue architecture where capillaries were predominantly in approximate cross‐section and CD31 staining of endothelial cells and capillaries was strongly positive.

Figure 7. Coronary arteries involved and percentage area stenosis for 48 decedents with coronary artery disease.

Shown are the numbers of decedents with coronary stenoses of the left main (A), left anterior descending (B), left circumflex (C), and right coronary arteries (D), and the extent of disease (percentage area stenosis) for 48 decedents (25 male and 23 female) with coronary artery disease. All decedents with coronary artery disease had ≥1 coronary artery with ≥75% area stenosis.

Figure 8. Cardiomyocyte width, total myocardial fibrosis, and arteriole perivascular fibrosis in decedents with and without coronary artery disease.

Shown are the cardiomyocyte width (A), total myocardial fibrosis (B), and arteriole perivascular fibrosis (C) for inner, middle, and outer regions of left ventricle sections from men and women without (NoCAD, open columns) and with CAD (closed columns). Data shown as means±SD, n=25 for men and women without CAD, and n=25 for men and n=23 for women with CAD. Mean differences (95% CI) between men and women and between decedents with and without CAD, were calculated for all regions, and for the inner, middle, and outer regions of each section; *P<0.05; **P<0.01; ***P<0.001. The higher total myocardial fibrosis (P=0.0010) and lower perivascular fibrosis (P=0.0039) in decedents with CAD were confirmed in a mutlivariable model with regions (inner, middle, outer) as the dependent variable and CAD, sex, and age category as independent variables. The higher perivascular fibrosis in men than women was also confirmed in the multivariable model (P=0.049). CAD indicates coronary artery disease.

Figure 9. Arteriole density, minimal arteriole diameter, and arteriole wall area/circumference ratio in decedents with and without coronary artery disease.

Shown are arteriole density (A), minimal arteriole diameter (B), and arteriole wall area/circumference ratio (C) for inner, middle, and outer regions of left ventricle sections from men and women without (NoCAD, open columns) and with CAD (closed columns). Data shown as means±SD, n=25 for both men and women without CAD and n=25 for men and n=23 for women with CAD. Mean differences (95% CI) between men and women and between decedents with and without CAD were calculated for all regions and for the inner, middle, and outer regions of each section; *P<0.05; **P<0.01; ***P<0.001. The lower arteriole density (P=0.0012) and higher median arteriole diameter (P<0.0001) in decedents with CAD were confirmed in a multivariable model with regions (inner, middle, outer) as the dependent variable and CAD, sex, and age category as independent variables. However, minimal arteriole diameter was not different between men and women in the multivariable model. CAD indicates coronary artery disease.

Figure 10. Correlations between median minimal arteriole diameter and arteriole density for inner, middle, and outer regions of each left ventricle section for decedents without and with coronary artery disease.

Shown are correlations between median minimal arteriole diameter and arteriole density for inner (A), middle (B), and outer (C) regions of left ventricle sections from decedents (male and female) without (NoCAD, left panels) and with CAD (right panels). Each dot represents a decedent; n=50 for decedents (25 male and 25 female) without CAD and n=48 for decedents (25 male and 23 female) with CAD. Also shown are Pearson correlation coefficients with 95% CI, all P values <0.0001. CAD indicates coronary artery disease.

Figure 11. Arteriole density for inner, middle and outer regions of each section of left ventricle, stratified according to minimal arteriole diameter, for decedents without and with coronary artery disease.

Shown are arteriole density for inner (A), middle (B), and outer (C) regions of left ventricle sections from decedents (male and female) without (NoCAD, open columns) and with CAD (closed columns). Data shown as means±SD, n=50 for decedents (25 male and 25 female) without CAD and n=48 for decedents (25 male and 23 female) with CAD. Mean differences (95% CI) in arteriole density between decedents with and without CAD were calculated for all diameters and for each diameter category; *P<0.05; **P<0.01; ***P<0.001. CAD indicates coronary artery disease.

Figure 12. Capillary length density, diffusion radius, and diffusion radius/cardiomyocyte width ratio in decedents with and without coronary artery disease.

Shown are capillary length density (A), diffusion radius (B), and diffusion radius/cardiomyocyte width ratio (C) for inner, middle, and outer regions of left ventricle sections from men and women without (NoCAD, open columns) and with CAD (closed columns). Data shown as means±SD, n=24 for both men and women without CAD and n=22 for men and n=20 for women with CAD. Mean differences (95% CI) between men and women and between decedents with and without CAD were calculated for all regions and for the inner, middle, and outer regions of each section; *P<0.05; **P<0.01; ***P<0.001. The lower capillary length density (P<0.0001), higher diffusion radius (P<0.0001), and higher diffusion radius/cardiomyocyte width ratio (P=0.0075) in decedents with CAD were confirmed in a multivariable model with regions (inner, middle, outer) as the dependent variable, and CAD, sex, and age category as independent variables. CAD indicates coronary artery disease.

Figure 13. Threshold extent of coronary artery disease for relationship with total myocardial fibrosis, arteriole density, and capillary length density.

Shown are total myocardial fibrosis (A), arteriole density (B) and capillary length density (C) for decedents with different extent of CAD. All decedents with CAD had ≥1 coronary artery with ≥75% area stenosis. Data shown as means±SD for all regions (inner, middle, and outer) for men and women combined, number of decedents (n) shown within the columns. Mean differences (95% CI) were calculated for the comparison of decedents with no CAD (NoCAD) and decedents with mild/minimal CAD (Mild/Minimal CAD, left panels); additionally, mean differences (95% CI) were calculated for the comparison of decedents with no or mild/minimal CAD (No/Mild/Minimal CAD) and decedents with either 1, 2, or 3 coronary arteries with ≥50% area stenosis (right panels). Comparisons of No/Mild/Minimal CAD with CAD (1, 2, or 3 coronary arteries with ≥50% area stenosis) were performed without adjustment for multiple comparisons. *P<0.05; **P<0.01; ***P<0.001. CAD indicates coronary artery disease.

Figure 14. Correlations between arteriole density and total myocardial fibrosis in decedents with and without coronary artery disease.

Shown are correlations between arteriole density and total myocardial fibrosis for the inner (A), middle (B), and outer (C) regions of left ventricle sections from men and women combined without (NoCAD, left panels) and with CAD (right panels). Each dot represents a decedent; n=50 (25 male, 25 female) for NoCAD and n=48 (25 male, 23 female) for CAD. Also shown are Pearson correlation coefficients with 95% CI, all P values <0.05. CAD indicates coronary artery disease.

Figure 15. Correlations between capillary length density and cardiomyocyte width, and between arteriole density and cardiomyocyte width, for inner, middle, and outer regions of each left ventricle section for decedents without and with coronary artery disease.

Shown are correlations between capillary length density and cardiomyocyte width for left ventricle sections from decedents without (A) and with CAD (B) and between arteriole density and cardiomyocyte width for left ventricle sections from decedents without (C) and with CAD (D). Each dot represents an inner, middle, or outer region of each left ventricle section; n=144 (inner, middle, or outer regions for 24 men and 24 women) for capillary length density for decedents without CAD; n=126 (inner, middle, or outer regions for 22 men and 20 women) for capillary length density for decedents with CAD; n=150 (inner, middle, or outer regions for 25 men and 25 women) for arteriole density for decedents without CAD; n=144 (inner, middle, or outer regions for 25 men and 23 women) for arteriole density for decedents with CAD. Also shown are Pearson correlation coefficients with 95% CI, all P values <0.01. CAD indicates coronary artery disease.