Dysglycemia and the Density of the Coronary Vasa Vasorum

Diabetes promotes retinal, renal, neurologic, vascular, cardiac, and other organ damage. Although many reasons for this have been suggested, dysglycemia-mediated vasculopenia due to capillary damage in various tissues would provide an overarching explanation. In the walls of medium to large arteries, these damaged capillary beds would be the vasa vasorum (1), a possibility supported by animal models (2) and in vivo measurements in humans (3,4).

If the above hypothesis is true, the density of the vasa vasorum should be reduced in the coronary arteries of people with diabetes. This cross-sectional autopsy study was therefore conducted to compare the coronary artery vasa vasorum in individuals with normal versus high HbA_1c at the time of death and to assess the relationship between HbA_1c and vasa vasorum density.

Deceased individuals undergoing forensic autopsy whose families consented to a full or limited postmortem examination were included if they were aged 50 or older, their coronary arteries could be sectioned and stained, a postmortem blood sample was available, and the autopsy occurred during the pathologist’s (V.N.) working hours. Age, sex, history of diabetes, and cause of death were collected, and a postmortem whole-blood sample was sent for HbA_1c. The study was approved by the Research Ethics Board at Hamilton Health Sciences and McMaster University.

Left anterior descending coronary arteries were formalin fixed, decalcified, and sectioned. The sections were embedded in paraffin, and the proximal portions were sectioned for future immunohistochemical staining. The slides were then heat-fixed for 4 min, and paraffin was removed using xylene solution for 5 min. Slides were passed through serial concentrations of ethanol (100% for 15 min, 70% for 6 min, and 50% for 6 min) and then rehydrated in distilled H₂O for 20 min.

Antigen unmasking solution (H-3300; Vector Laboratories) was boiled for 17 min, and the sections were then placed in the boiling solution for 3 min. After cooling for 20 min, they were 1) blocked and permeabilized in 10% goat serum (Vector Laboratories) + 0.05% Triton-X 100 in PBS (pH 7.4) at room temperature for 45 min, 2) rinsed in PBS (pH 7.4), and 3) incubated overnight at 4°C with rabbit anti-human von Willebrand factor polyclonal antibody (DAKO) that was diluted 1:200 in 10% goat serum + 0.05% Triton-X 100 in PBS (pH 7.4). The sections were then transferred to 37°C for 1 h, washed four times (for 5 min each) with PBS (pH 7.4), and incubated with the secondary antibody (goat anti-rabbit IgG) that was conjugated to Alexa Fluor 594 (Life Technologies), diluted 1:250 in PBS (pH 7.4), and protected from light for 2 h at room temperature. The sections were then washed four times (for 5 min each) in PBS (pH 7.4) and counterstained for 2 min with the nuclear stain DAPI (Invitrogen), diluted 1:5,000 in PBS (pH 7.4). After being washed four times (for 5 min each) in PBS (pH 7.4), they were mounted with the Fluorescent Aqueous Mounting Medium (Sigma-Aldrich). To control for nonspecific fluorescence, preimmune IgG control was used instead of the primary antibody. Slides were stored in the dark at 4°C until imaged.

All specimens were interpreted without knowledge of diabetes status or the HbA_1c. The vasa vasorum were imaged using a fluorescent microscope (Nikon Eclipse T), and three sections were stained per individual. Microvessel density (vessels/mm²) was quantified for each section using Nikon Instruments Software (NIS)-Elements, and the mean density was calculated for each individual. Staining exceeding a set fluorescence intensity was counted in each of the intimal-medial and adventitial layers.

Because the distribution of the density of coronary vasa vasorum was unknown, no formal sample size calculations were done. The study was designed to continue until 20 individuals with diabetes were included, as defined above. It stopped after 19 individuals with diabetes were included due to relocation of the investigator pathologist.

HbA_1c was analyzed on the Bio-Rad Variant II Instrument in the Hamilton Health Sciences National Glycohemoglobin Standardization Program–certified Clinical Research Laboratory. Diabetes was diagnosed based on a known diagnosis or a postmortem HbA_1c level ≥6.5% (48 mmol/mol). Groups were compared using ordinary one-way ANOVA, and multiple linear regression estimated the age- and sex-adjusted effect of HbA_1c on capillary density within the various layers. Statistical analyses were done using GraphPad Prism and SAS 9.4 software.

The coronary arteries of 57 individuals were analyzed (Supplementary Table 1). Their mean age was 62.5 years, 36 (63%) were men, their mean HbA_1c was 6.25% (45 mmol/mol), and cancer or coronary artery disease was listed as the cause of death in 24 (42.1%). The mean (SD) capillary density in the adventitia, intima-media, and the entire vessel wall was 33.2 (13.7), 18.5 (8.3), and 25.6 (9.8) vessels/mm², respectively.

At the time of death, 22 individuals (38.6%) were normoglycemic, 17 (29.8%) had an HbA_1c level consistent with prediabetes (5.7–6.4% or 39–46 mmol/mol), and 18 (31.6%) had a prior diagnosis of diabetes or an HbA_1c level ≥6.5% (48 mmol/mol). Within both the adventitia and the entire vessel wall (Supplementary Fig. 1), the capillary density was significantly lower in people with diabetes versus each of the other groups (P < 0.005). In a multiple linear regression model adjusted for age and sex, a 1% higher HbA_1c was associated with a 3.2 vessels/mm² (95% CI 0.6, 5.8) lower adventitial capillary density (P = 0.017) and a 2.3 vessels/mm² (95% CI 0.6, 4.2) lower vessel wall capillary density (Fig. 1).

This study was designed to test the previously published hypothesis that diabetes-related coronary artery disease is due to dysglycemia-mediated capillary damage in the vasa vasorum of arteries (1). This hypothesis is supported by our findings that people with diabetes had a lower density of the coronary vasa vasorum than those without diabetes and that the higher the HbA_1c, the lower the density of these vessels in the adventitia and entire vessel wall. The HbA_1c range defining prediabetes is much smaller than that characterizing diabetes, which likely accounts for the absence of a difference between those with normoglycemia and prediabetes.

Several Mendelian randomization studies (5–7), epidemiologic evidence (8), and some trials have demonstrated that dysglycemia is causally linked to ischemic heart disease (9) and other consequences of diabetes (10). These findings suggest that one mechanism for this relationship could be glucose-mediated vasculopenia in the adventitial capillary network of the coronary arteries. This may promote hypoxia-mediated damage to the vessel wall and/or inflammation, angiogenesis, and subsequent plaque formation and rupture (11). If the link between capillary integrity and diabetes-related disease in other organs is confirmed by in vivo imaging of organ-specific capillary beds and measurement of tissue perfusion (12), drugs that block the effect of dysglycemia on capillaries may also reduce cardiovascular and other diabetes consequences.

Unique strengths of this study are that these findings arose from a study explicitly designed to test a previously articulated hypothesis (1) and are supported by evidence from animal experiments (2) and the analysis of tissue (rather than ultrasound or other images) from unselected individuals. However, the cross-sectional design and limited data mean that the possibility that findings are a consequence of coronary disease cannot be excluded. Nevertheless, these findings extend the list of capillary beds damaged by diabetes from the eyes and kidneys to the vasa vasorum and support the possibility that dysglycemia may promote a similar vasculopenic process in capillary beds in the heart muscle, skeletal muscle, brain, liver, and other organs.

Funding. This project was supported by an operating grant from the Heart and Stroke Foundation of Canada (G-17-0017029).

Duality of Interest. H.C.G. holds the McMaster-Sanofi Population Health Institute Chair in Diabetes Research and Care and has received research grant support from AstraZeneca, Eli Lilly, Merck, and Sanofi; honoraria for speaking from AstraZeneca, Boehringer Ingelheim, Eli Lilly, Novo Nordisk, and Sanofi; and consulting fees from Abbott, AstraZeneca, Boehringer Ingelheim, Eli Lilly, Merck, Novo Nordisk, Janssen, and Sanofi. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. H.C.G. wrote the first draft of the manuscript. V.N., R.C., H.S., and G.W. reviewed and edited the manuscript and acquired and analyzed the samples. H.C.G., V.N., and G.W. developed the protocol and analyzed the data. H.C.G. was responsible for the study design and the decision to submit and publish the manuscript. H.C.G. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.