Am J Physiol Heart Circ Physiol 285: H2019–H2026, 2003.
First published July 10, 2003; 10.1152/ajpheart.00399.2003.
Impact of coronary vasa vasorum functional structure
on coronary vessel wall perfusion distribution
M. Gössl,1 N. M. Malyar,1 M. Rosol,2 P. E. Beighley,1 and E. L. Ritman1
1
Physiological Imaging Research Laboratory, Department of Physiology and Biophysics, Mayo Clinic,
Rochester, Minnesota 55905; and 2Childrens Hospital Los Angeles, Los Angeles, California 90027
Submitted 22 May 2003; accepted in final form 2 July 2003
VASA VASORUM HAVE BEEN DESCRIBED as a “network” or
“plexus” of microvessels located in the vessel wall of
arteries and veins responsible for nutrient supply to
and drainage from the vessel wall (2, 9, 21, 24, 25). A
recent study by our group (5) using microcomputerized
tomography (micro-CT) showed, however, that the general structure of coronary vasa vasorum is “treelike,”
consistent with endarteries. Furthermore, we could
show that despite the constrained topology of the space
in which they proliferate, the hemodynamic characteristics of coronary vasa vasorum are surprisingly similar to that of the vasculature in general. Although
several investigators (6, 14, 17) have shown that blood
flow in vasa vasorum can be regulated by different
vasoactive agents such as adenosine, endothelin-1, or
acetylcholine, the question as to whether vasa vasorum
function as networks or endarteries has not been resolved. Both structures (network or endartery) have
different implications as to the perfusion of the coronary vessel wall. A network can maintain perfusion
more easily by virtue of shunting and retrograde perfusion, whereas the occlusion of a proximal endartery
branch will stop perfusion of all distal branches. However, even if there is an anatomic plexus of vasa vasorum, local occlusion of that network by the compressive
forces inside the coronary artery wall could make the
vasa vasorum functionally endarteries.
Hence, the present study was designed to investigate
the functional branching structure of vasa vasorum
through the impact on their perfusion in two different
conditions of coronary artery wall compressive forces
relative to the perfusion pressure within the vasa vasorum. The first condition studied was experimental
microembolization. The occlusion of proximal segments
of a vasa vasorum structure may indicate which functional structure is predominant. If it was the network
or plexus, then the occlusion of a single proximal segment should not have much impact on the perfusion of
more distally located vasa vasorum segments. If, however, the vasa vasorum were endarteries, the impact
would be significant, because none of the more distal
vasa vasorum segments would be perfused. The second
condition studied was a change of coronary artery
intramural pressure. The intraluminal coronary artery
pressure produces compressive forces inside the coronary artery wall [Lamé’s Law (3)] and thus may occlude those distal vasa vasorum segments that have
Address for reprint requests and other correspondence: E. L.
Ritman, Physiological Imaging Research Laboratory, Dept. of Physiology and Biophysics, Mayo Clinic, 200 First St. SW, Rochester, MN
55905 (E-mail: [email protected]).
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
intraluminal coronary artery pressure; microembolization;
microcomputerized tomography
http://www.ajpheart.org
0363-6135/03 $5.00 Copyright © 2003 the American Physiological Society
H2019
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on June 14, 2017
Gössl, M., N. M. Malyar, M. Rosol, P. E. Beighley, and
E. L. Ritman. Impact of coronary vasa vasorum functional
structure on coronary vessel wall perfusion distribution. Am
J Physiol Heart Circ Physiol 285: H2019–H2026, 2003. First
published July 10, 2003; 10.1152/ajpheart.00399.2003.—
Noncoronary vasa vasorum have been described as networks
of microvessels in the wall of arteries and veins. However, we
have shown, using microcomputerized tomography (microCT) imaging methods, that porcine coronary vasa vasorum
have a tree-like branching structure similar to the vasculature in general. In this study, we elucidate functional aspects
of coronary vasa vasorum perfusion territories. Three pig
hearts were injected with radiopaque Microfil via the coronary sinus to fill the left anterior descending coronary arteries (LADs) retrogradely at atmospheric pressure. In three
other hearts, LADs were injected antegradely at 100-mmHg
pressure via the left main carotid artery. Additionally, six
LADs were injected in vivo with a suspension of 100- or
300-␮m-diameter microspheres before harvesting of the
hearts and injection of the LADs with Microfil. All harvested
LADs were scanned intact with micro-CT (20 ␮m cubic voxels). The spatial density of vasa vasorum (no. of vasa/mm2)
was measured in 20-␮m-thick cross sections (at 0.4-mm intervals). Retrogradely injected LADs showed high and uniformly distributed vasa vasorum densities in the adventitia
(means ⫾ SE; 5.38 ⫾ 0.09 vs. 3.58 ⫾ 0.1 vasa/mm2 in
antegradely prepared LADs; P ⬍ 0.001). Antegradely prepared LADs showed patchy distributed, low-vasa-vasorumdensity territories especially on the myocardial side of the
coronary artery wall (epicardial density: 4.29 ⫾ 0.13 vasa/
mm2 vs. myocardial density: 2.80 ⫾ 0.1 vasa/mm2, P ⬍
0.001). Microembolization reduced vasa vasorum densities
significantly (100-␮m-diameter microspheres: 3.26 ⫾ 0.07
vasa/mm2, P ⬍ 0.05; 300-␮m-diameter microspheres: 2.66 ⫾
0.07 vasa/mm2, P ⬍ 0.001 vs. antegrade controls) and increased the size of low-vasa-vasorum-density territories. We
conclude that coronary vasa vasorum are functional endarteries not connected via a plexus. This characteristic may
have a significant impact on the spatial distribution of perfusion and drainage of the coronary vessel wall.
H2020
FUNCTIONAL STRUCTURE OF CORONARY VASA VASORUM
reduced intraluminal pressure due to resistive forces
[Poiseuille’s Law (22)] compared with the compressive
forces surrounding them.
METHODS
Animal Studies and Specimen Acquisition
Microscopic 3-D CT Reconstruction
Specimens up to 10 cm long were scanned intact by a
micro-CT system consisting of a spectroscopy X-ray tube, a
fluorescent crystal plate, a microscopic objective, and a
charge-coupled device (CCD) camera (8). To preserve the
connectivity of the vasa vasorum, the arteries were scanned
in 2-cm increments along the coronary artery luminal axis
(i.e., without cutting the 6- to 10-cm-long coronary artery into
pieces). The resulting 3-D images were displayed using image analysis software (Analyze 4.0, Biomedical Imaging Resource, Mayo Clinic, Rochester, MN). Computer-generated
displays were generated at different angles of view and gray
scale values of voxels. For this study, the micro-CT scanner
was configured so that the dimension of the cubic voxels was
20 ␮m (16-bit gray scale).
Delineating the Coronary Vessel Wall and Quantitative
Morphometric Analyses
Fig. 1. Schematic representation of the microembolization experiment. As shown in earlier experiments (5), porcine coronary vasa
vasorum have a treelike branching structure such that proximal
occlusion by microemboli prevents perfusion of distal vasa vasorum
tree segments (indicated here by the dark-colored vasa vasorum
interna, which is embolized at its origin from the epicardial coronary
artery lumen). LAD, left anterior descending coronary artery; LCX,
left circumflex artery; micro-CT, microcomputerized tomography.
AJP-Heart Circ Physiol • VOL
After the arteries had been scanned with micro-CT, we
obtained histological cross sections (Lawson’s Elastic Van
Gieson staining) from all arteries (1 cross section every 2 cm)
and took digitized microscopic photographs. On these images, we measured the circumferences (C) of 1) the lumen, 2)
the interface of media and adventitia (i.e., the outer elastic
membrane), and 3) the outer surface of the adventitia using
Analyze software. Next, we calculated the radius (r ⫽ C/2␲)
for each of these three perimeters. The ratio of the radii
(denominator was the lumen radius) gave us the scale factor
285 • NOVEMBER 2003 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on June 14, 2017
All animal studies were approved by the Mayo Foundation’s Institutional Animal Care and Use Committee. Twelve
female, juvenile, domestic, cross-bred swine (31–42 kg body
wt) were fed normal laboratory chow for 3 mo. Six animals
were divided into two equal groups: group a received in vivo
injections of suspensions of 300-␮m-diameter, nonradioactive
microspheres (2 ⫻ 102 microspheres), and group b received
injections of 100-␮m-diameter microspheres (5 ⫻ 103 microspheres). Swine were anesthetized, intubated, and ventilated
with room air as described previously (10). Under sterile
conditions, the neck was dissected on the left side, and the
left jugular vein and carotid artery were isolated. A catheter
was positioned in the left anterior descending coronary artery (LAD) between the first and second septal branch (monitored by fluoroscopy). Finally, intracoronary injections of the
different amounts of microspheres were performed (Fig. 1).
Microspheres were prepared as described previously (16).
Electrodes on the limbs were used to monitor the electrocardiogram.
The results of our previous study (5) showed that there are
two types of arterial vasorum in the porcine coronary artery
wall: vasa vasorum externa (VVE) and vasa vasorum interna
(VVI). The average diameter of a major LAD branch that
leads to a VVE origin is 0.191 ⫾ 0.147 mm, and the average
diameter of VVI origin is 0.054 ⫾ 0.063 mm (5). Accordingly,
we chose 100- and 300-␮m-diameter microspheres for this
study.
An additional six swine served as controls. Microembolized
and control animals were euthanized with a weight-adjusted
intravenous injection of Sleepaway (Fort Dodge Laboratories;
Fort Dodge, IA). After the heart was harvested, we used two
different preparation techniques as follows.
Preparation 1. In the six microembolized and three control
hearts, we cannulated the LAD ostium with 2-mm-diameter
plastic cannulas. To clear the coronary network of blood, we
then infused 60 ml of heparinized saline (0.9% sodium chloride with 10,000 units heparin) via an automated injector at
a nominal pressure of 100 mmHg. Next, we infused a radiopaque, lead-containing, liquid, low-viscosity polymer (Microfil MV-122, Flow Tech; Carver, MA). This infusion (at 100
mmHg) was continued until the compound flowed freely from
the coronary sinus.
Preparation 2. In the three remaining control hearts, we
infused Microfil via the coronary sinus so as to fill the LAD
retrogradely, but without the high intra-arterial pressure.
This involved filling the ventricles and atria with low-melting
point paraffin [to prevent leakage of Microfil through arteriovenous shunts like the Thebesian veins (23)] and suturing
all openings that were produced by the harvesting process to
maintain pressure in the venous system. By infusing the
polymer retrogradely, the low pressure (open to atmosphere)
inside the LAD lumen prevented the compression of vasa
vasorum within the coronary vessel wall.
The hearts were immersed in 10% neutral buffered formalin and placed in refrigeration at 4°C overnight to allow
polymerization of the compound. On the following day, the
intact epicardial coronary arteries and concomitant veins
were removed from the heart with a margin of at least 1 cm
on each side of the artery, well outside the adventitia. The
specimens were then placed in a 75% alcohol solution. At
15-min intervals, the alcohol concentration was changed to
85%, 95%, and absolute alcohol to dehydrate the tissue. The
coronaries were encased in paraffin wax for micro-CT scanning and three-dimensional (3-D) reconstruction.
H2021
FUNCTIONAL STRUCTURE OF CORONARY VASA VASORUM
Subdivision of the Vessel Wall into Epicardial and
Myocardial Quadrants
To calculate the distribution of vasa vasorum densities
throughout the vessel wall, we subdivided the delineated
vessel wall into four quadrants [epicardial left (Epi-L), epicardial right (Epi-R), myocardial left (Myo-L), and myocardial right (Myo-R)]. This subdivision was done in every cross
section analyzed (see illustration for subdivision in Fig. 4).
Statistical Analysis
All data are presented as means ⫾ SE for all arteries.
One-way ANOVA, followed by a Tukey-Kramer post hoc test
with correction for multiple comparisons, was used to identify the statistical differences among groups. Individual
group comparisons were performed by an unpaired Student
t-test. P ⬍ 0.05 was considered significant in all analyses.
RESULTS
Densities and Spatial Distribution of Perfused
Coronary Vasa Vasorum
Impact of microembolization. In vivo microembolization with 100- or 300-␮m-diameter microspheres decreased the spatial density of coronary vasa vasorum
significantly (Table 1 and Fig. 2). The overall decrease
in vasa vasorum density after 300-␮m-microsphere
microembolization was significantly higher than that
with the 100-␮m-diameter microsphere microembolization (Table 1 and Figs. 2 and 3). Figure 3 shows the
impact of the different-sized microspheres on each of
the four vessel wall quadrants alone (subdivision of
coronary vessel wall in quadrants is illustrated in Fig.
4). Microembolization with 300-␮m-diameter microspheres decreased the density of vasa vasorum in all
four quadrants of the vessel wall significantly. Microembolization with 100-␮m-diameter microspheres decreased only the vasa vasorum density in the epicardial quadrants significantly.
Furthermore, Fig. 3 shows that in antegradely
(physiologically) prepared LADs, the epicardial side of
the vessel wall (Epi-L plus Epi-R) showed consistently
higher vasa vasorum densities than the myocardial
side (Myo-L plus Myo-R, P ⬍ 0.001). A subanalysis in
these arteries (Fig. 5) with exact differentiation of
venous and arterial vasa vasorum demonstrated that
only venous vasa vasorum showed significantly higher
spatial densities on the epicardial than on the myocardial side of the coronary artery wall. Arterial vasa
vasorum, however, were distributed equally between
the two sides of the coronary vessel wall.
Figure 4 shows the displays of spatial vasa vasorum
densities in all four quadrants of the coronary vessel
wall, referenced to the endothelial surface. The spatial
distribution of vasa vasorum densities in each vessel
wall quadrant along the coronary artery is encoded in
gray scale. Control LADs infused antegradely at a
physiological pressure of 100 mmHg showed a patchy
distribution of low-vasa-vasorum-density territories
(indicated by black areas). Consistent with the abovementioned average vasa vasorum densities (Fig. 3),
these low-density territories were preferentially located on the myocardial side (Myo-L and Myo-R) of the
arterial wall. In addition, as shown in Fig. 4 with
Table 1. Quantitative results of computerized digital micro-CT image analysis
Control Segments
n
Lumen radius, mm
Vessel wall area, mm2
“Spatial” density of vasa vasorum
(entire coronary vessel wall, vasa/mm2)
Vasa vasorum diameter, mm
Embolized Segments
Antegrade
injections
Retrograde
injections
100-␮m-diameter
Microspheres
300-␮m-diameter
Microspheres
364
0.85 ⫾ 0.01
2.87 ⫾ 0.06
429
0.81 ⫾ 0.02
3.22 ⫾ 0.07†
355
1.00 ⫾ 0.01*
3.51 ⫾ 0.07*
363
0.80 ⫾ 0.01†
2.76 ⫾ 0.07
3.58 ⫾ 0.1
0.056 ⫾ 0.001
5.38 ⫾ 0.09*
0.069 ⫾ 0.001‡
3.26 ⫾ 0.07‡
0.051 ⫾ 0.001
2.66 ⫾ 0.07*§
0.053 ⫾ 0.001
Data are means ⫾ SE; n, no. of cross sections analyzed. micro-CT, microcomputed tomography. * P ⬍ 0.001, † P ⬍ 0.01, and ‡ P ⬍ 0.05 vs.
control antegrade injected segments; § P ⬍ 0.01 vs. 100-␮m-diameter microsphere-embolized segments.
AJP-Heart Circ Physiol • VOL
285 • NOVEMBER 2003 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on June 14, 2017
for the transfer of these radii onto the corresponding micro-CT cross section. We used matching anatomic markers in
both the histological section and micro-CT cross section to
assure the correct correspondence of the micro-CT image to
the histological image. For those micro-CT cross sections
between the histological sections, we used the subtle intensity difference in the micro-CT image (appearing as a “halo”),
which matched well with the calculated interface of the
adventitia and the surrounding fat tissue. This finding provided us with the opportunity to determine and measure the
coronary artery wall (including the intima ⫹ media and
adventitia) in every single micro-CT cross section we analyzed without obtaining histology for each of them (5).
The spatial “density” of the vasa in the wall was obtained
by manually counting the number of vasa per 20-␮m-thick
cross-section divided by the total cross-sectional area (no. of
vasa/mm2) of vessel wall depicted in the tomographic image.
That way, vasa vasorum were counted in cross sections
spaced at 0.4-mm intervals along the LAD segments distal to
the microsphere injection site and in corresponding control
segments. Branching points were excluded from the analysis.
In summary, a total of 1,511 cross sections (i.e., 137 ⫾ 30
cross sections/specimen) was included in the analysis.
The differentiation of arterial and venous vasa vasorum
was done by using the 3-D information of the micro-CT data
set as described earlier (5). In brief, if the vasa vasorum could
be connected to the main lumen (VVI) or to a major branch
(VVE), they were considered arterial. In contrast, if they
could be connected to a concomitant vein, they were considered venous vasa vasorum.
H2022
FUNCTIONAL STRUCTURE OF CORONARY VASA VASORUM
Fig. 2. Three volume-rendered micro-CT images of a
normal LAD (left), a LAD that was embolized with
100-␮m-diameter microspheres (␮sph; middle), and a
LAD that was embolized with 300-␮m-diameter microspheres (right). Note the decrease in vasa vasorum in
the middle and the more severe decrease in the right,
where some parts of the LAD don’t show any vasa
vasorum.
vessel wall quadrants (Fig. 3 and Table 1) and there
was no significant difference between the epicardial
and myocardial quadrants (5.5 ⫾ 0.3 vs. 5.2 ⫾ 0.3
vasa/mm2, not significant). The vasa vasorum diameter
increased ⬃23% compared with antegradely prepared
coronary arteries (Table 1).
DISCUSSION
3-D micro-CT imaging offers the particular important capability to conveniently analyze large 3-D volumes of coronary arteries without the need for physical
sectioning, thus preserving the continuity of the vascular structures. It allows quantification of the vasa
vasorum without the need to prepare hundreds of timeand cost-intensive histological sections (not to mention
the technical difficulties in preventing distortion, accurate section-to-section registration, and visualization
of the presence of vasa vasorum in histology). Hence,
Fig. 3. Spatial vasa vasorum densities within
the four quadrants of the vessel wall of control
LADs (with antegrade and retrograde infusion
of Microfil) and LADs that were embolized with
100- and 300-␮m-diameter microspheres. Control LADs (antegrade preparation) show a significant difference in vasa vasorum spatial densities between epicardial and myocardial quadrants. This difference remains significant in the
100- and 300-␮m-diameter microsphere-embolized LADs. Furthermore, microembolization
with 100- and 300-␮m-diameter microspheres
decreases the vasa vasorum density significantly, with the highest impact by 300-␮mdiameter microspheres. The retrogradely infused control LADs show significantly higher
vasa vasorum densities in all quadrants and no
significant difference between epicardial and
myocardial quadrants. Epi-R and Epi-L, epicardial left and right quadrants; Myo-L and
Myo-R, myocardial left and right quadrants.
*P ⱕ 0.001, #P ⬍ 0.05, and †P ⬍ 0.01 vs. control
LADs and epicardial vs. myocardial quadrants;
‡P ⬍ 0.001 vs. 100-␮m-diameter microspheres.
AJP-Heart Circ Physiol • VOL
285 • NOVEMBER 2003 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on June 14, 2017
representative examples, microembolization with 100and 300-␮m-diameter microspheres increased the size
of those patchy distributed low-density territories significantly [antegrade controls: 4.99 ⫾ 1.78 mm2; 100␮m-diameter microspheres: 11.26 ⫾ 1.82 mm2 (P ⬍
0.05 vs. controls); 300-␮m-diameter microspheres:
18.09 ⫾ 1.94 mm2 (P ⬍ 0.001 vs. controls and P ⬍ 0.05
vs. 100-␮m-diameter microspheres)].
Impact of change in coronary artery intramural pressure. Figure 4 also shows that, in contrast to antegradely (i.e., with 100-mmHg intraluminal coronary
artery pressure) prepared LADs, retrogradely perfused
LADs (i.e., with intraluminal coronary artery pressure
open to atmosphere) showed fewer, relatively small,
patchy distributed low-vasa-vasorum-density territories [1.65 ⫾ 0.39 mm2 (P ⬍ 0.001 vs. antegrade controls)]. Consistently, under atmospheric intraluminal
pressure, the average spatial density of coronary vasa
vasorum was significantly higher in all coronary artery
FUNCTIONAL STRUCTURE OF CORONARY VASA VASORUM
H2023
we could precisely calculate vasa vasorum densities in
more than 1,500 digital cross sections and reference
them to the endothelial surface of the host vessel. In
the present study, the micro-CT scanner was configured to generate 20 ␮m/side cubic voxels so as to
maximize the intact volume that could be scanned
while maintaining a reasonable spatial resolution.
Consequently, the 20-␮m voxel size of our 3-D micro-CT data limited the diameter of vessels that could
be visualized, so that there are possibly smaller vasa
vasorum inside the media or adventitia that were not
resolvable in our micro-CT images.
Several investigators (2, 9, 21, 24, 25) have described
the vasa vasorum as a network or plexus of microvessels in the wall of arteries, although these results were
based on observations in the presence of vascular disAJP-Heart Circ Physiol • VOL
ease. The results of the present study in normal porcine coronary arteries, however, provide evidence that
coronary vasa vasorum are functional endarteries
rather than a functional network or plexus. An endartery, by definition, is not connected to other vessels’
perfusion territories, in contrast to a network or
plexus, where different vessels’ perfusion territories
may communicate, both antegrade and retrograde,
with each other. Hence, the occlusion of a segment that
is part of a network still allows perfusion of more
distally located segments by virtue of redirected flow in
the network. The compression or occlusion of a proximal segment of an endartery, however, leads to a stop
of perfusion of the more distal segments of the endartery. It is important to know which of the two possibilities is predominant in coronary vasa vasorum to un-
285 • NOVEMBER 2003 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on June 14, 2017
Fig. 4. Display of vasa vasorum densities (arterial and venous vasa vasorum in the vessel
wall of the LAD) in all four vessel wall quadrants, referenced to the endothelial surface,
shown in representative examples of an antegradely infused control LAD (data from 128
cross sections at 0.4-mm intervals), a 100␮m-diameter microsphere-embolized artery
(117 cross sections), a 300-␮m-diameter microsphere-embolized artery (136 cross sections), and a retrogradely infused control
LAD (101 cross sections). The densities of
vasa vasorum are encoded in grayscale. This
display shows the presence of a patchy distribution of low-vasa-vasorum-density territories (black) in the antegrade control. These
low-density territories enlarge after 100and 300-␮m-diameter microsphere injections
(⫹48% and ⫹61%, respectively). The retrograde control shows almost no low-vasa-vasorum-density territories (⫺92%).
H2024
FUNCTIONAL STRUCTURE OF CORONARY VASA VASORUM
Fig. 5. Spatial density of venous and arterial porcine coronary vasa vasorum.
Only venous vasa vasorum show a significant difference in density between the
myocardial and epicardial sides of the coronary vessel wall. Arterial vasa vasorum
are uniformly distributed. *P ⬍ 0.001.
AJP-Heart Circ Physiol • VOL
have two implications: 1) that the reduced compressive
forces within the arterial wall only compress the very
distal vasa vasorum branches or 2) that a few small
regions within the coronary vessel wall contain no vasa
vasorum at all.
Our results are based on static data, and this is a
possible limitation in that oscillating changes of vasa
vasorum intraluminal pressure may log the phase of
the compressive forces inside the coronary artery wall.
The high number of vasa vasorum filled under minimal
coronary arterial intraluminal pressure (i.e., retrograde filling of the LAD; Fig. 3) may indicate that
during certain time points within the cardiac cycle,
those vessels might be recruitable as well. Another
possible limitation may be seen in the number of animals studied, but because the intragroup variability
was statistically insignificant, we are confident that
our results are representative.
More evidence for vasa vasorum being functional
endarteries can be deducted from our microembolization experiments. Microembolization, by virtue of occlusion of proximal vasa vasorum segments, decreased
the number of perfused arterial vasa vasorum (Figs. 2
and 3) and increased the size of patchy distributed
low-vasa-vasorum-density territories in the coronary
artery wall (Fig. 4). If vasa vasorum form a functional
plexus, the occlusion of some proximal segments of the
vasa vasorum by microemboli should be compensated
for by redirected flow within the network. The probability for a high number of interconnecting branches in
a network being embolized and causing the same or
even higher impact on vessel wall perfusion is negligible and, moreover, not consistent with our earlier,
detailed analysis of the anatomic structure of porcine
coronary vasa vasorum (5). The number of vasa vasorum perfused (Fig. 3), and hence the size of low-vasavasorum-density territories (Fig. 4, black areas), is
conceivably dependent on the hierarchical location of
the occlusion within the tree structure, as indicated by
the use of different-sized microspheres in our experi-
285 • NOVEMBER 2003 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on June 14, 2017
derstand in how far disturbances in their structure can
impact on the perfusion of the coronary vessel wall.
We demonstrated a feasible method for the retrograde, minimal pressure filling of a coronary artery.
This method essentially eliminates the pressure gradient within the arterial wall. In conjunction with the
method of antegrade (physiological) filling, these two
experiments allowed us to analyze the impact of the
intraluminal coronary artery pressure alone and hence
the different intramural compressive forces (radial
stress) on the distribution and diameter of vasa vasorum inside the coronary artery vessel wall.
We deduced from the increase in vasa vasorum diameters (⬎23%) under atmospheric intraluminal coronary artery pressure that the compressive forces inside
the coronary artery wall have significant impact on
flow through the vasa vasorum. The results show,
furthermore, that a physiological coronary arterial intraluminal pressure (100 mmHg) by virtue of compressive forces in the arterial wall leads to a significant
decrease in the number of perfused vasa vasorum (Fig.
3) and a patchy distribution of low-vasa-vasorum-density territories (Fig. 4). If vasa vasorum are a network
or plexus, as postulated by several investigators, the
compression of segments of that network should be
compensated by redirected flow through the multiple
connections of the network. However, the patchy distribution of the low-density territories (Fig. 4) strongly
indicates that vasa vasorum are functional endarteries
with a treelike branching structure, consistent with
our earlier findings (5). Hence, the compression of
proximal segments of the treelike structure leads to
nonperfused, distal tree segments, creating patchy distributed low-vasa-vasorum-density territories. In contrast, under a minimal (atmospheric) coronary arterial
intraluminal pressure (i.e., minimal intramural compressive forces after retrograde perfusion), there were
only a few, small low-density territories in the coronary
artery wall (Fig. 4), and the number of perfused vasa
vasorum is significantly higher (Fig. 3). These findings
FUNCTIONAL STRUCTURE OF CORONARY VASA VASORUM
We thank Julie M. Patterson for help with the figures.
DISCLOSURES
This study was supported in part by National Institutes of Health
Grants HL-65342 and EB-000305 and by the Mayo Clinic.
AJP-Heart Circ Physiol • VOL
REFERENCES
1. Barker SG, Talbert A, Cottam S, Baskerville PA, and Martin JF. Arterial intimal hyperplasia after occlusion of the adventitial vasa vasorum in the pig. Arterioscler Thromb 13: 70–
77, 1993.
2. Bo WJ, McKinney WM, and Bowden RL. The origin and
distribution of vasa vasorum at the bifurcation of the common
carotid artery with atherosclerosis. Stroke 20: 1484–1487,
1989.
3. Den Hartog JP. Strength of Materials. New York: Dover, 1949.
4. Dorge H, Neumann T, Behrends M, Skyschally A, Schulz
R, Kasper C, Erbel R, and Heusch G. Perfusion-contraction
mismatch with coronary microvascular obstruction: role of inflammation. Am J Physiol Heart Circ Physiol 279: H2587–
H2592, 2000.
5. Gössl M, Rosol M, Malyar NM, Fitzpatrick LA, Beighley
PE, Zamir M, and Ritman EL. Functional and hemodynamic
characteristics of vasa vasorum in the walls of porcine coronary
arteries. Anat Rec 272A: 526–537, 2003.
6. Heistad DD, Marcus ML, Law EG, Armstrong ML,
Ehrhardt JC, and Abboud FM. Regulation of blood flow to the
aortic media in dogs. J Clin Invest 62: 133–140, 1978.
7. Jeremias A, Huegel H, Lee DP, Hassan A, Wolf A, Yeung
AC, Yock PG, and Fitzgerald PJ. Spatial orientation of atherosclerotic plaque in non-branching coronary artery segments.
Atherosclerosis 152: 209–215, 2000.
8. Jorgensen SM, Demirkaya O, and Ritman EL. Three-dimensional imaging of vasculature and parenchyma in intact rodent
organs with X-ray micro-CT. Am J Physiol Heart Circ Physiol
275: H1103–H1114, 1998.
9. Kwon HM, Sangiorgi G, Ritman EL, McKenna C, Holmes
DR Jr, Schwartz RS, and Lerman A. Enhanced coronary vasa
vasorum neovascularization in experimental hypercholesterolemia. J Clin Invest 101: 1551–1556, 1998.
10. Liu YH, Bahn RC, and Ritman EL. Microvascular blood
volume-to-flow relationships in porcine heart wall: whole body
CT evaluation in vivo. Am J Physiol Heart Circ Physiol 269:
H1820–H1826, 1995.
11. McGill HC Jr, McMahan CA, Herderick EE, Zieske AW,
Malcom GT, Tracy RE, and Strong JP. Obesity accelerates
the progression of coronary atherosclerosis in young men. Circulation 105: 2712–2718, 2002.
12. Nakata Y and Kamiya K. An experimental study on the vascular lesions caused by obstruction of the vaso vasorum. II.
Special consideration on the deposition of fat into vascular wall.
Jpn Circ J 34: 1029–1034, 1970.
13. Nakata Y, Shionoya S, Matsubara J, and Shinjo K. An
experimental study on the vascular lesions caused by disturbance of the vasa vasorum. 3. Influence of obstruction of the
venous side of the vasa vasorum and the periaortic vein. Jpn
Circ J 36: 945–951, 1972.
14. Ohhira A and Ohhashi T. Effects of aortic pressure and
vasoactive agents on the vascular resistance of the vasa vasorum in canine isolated thoracic aorta. J Physiol 453: 233–245,
1992.
15. Ojha M, Leask RL, Butany J, and Johnston KW. Distribution of intimal and medial thickening in the human right coronary artery: a study of 17 RCAs. Atherosclerosis 158: 147–153,
2001.
16. Schmermund A, Lerman LO, Rumberger JA, Lund PE,
Pfeifer EA, Sheedy PF II, and Ritman EL. Effects of acute
and chronic angiotensin receptor blockade on myocardial vascular blood volume and perfusion in a pig model of coronary
microembolization. Am J Hypertens 13: 827–837, 2000.
17. Scotland R, Vallance P, and Ahluwalia A. On the regulation
of tone in vasa vasorum. Cardiovasc Res 41: 237–245, 1999.
18. Skalen K, Gustafsson M, Rydberg EK, Hulten LM, Wiklund O, Innerarity TL, and Boren J. Subendothelial retention of atherogenic lipoproteins in early atherosclerosis. Nature
417: 750–754, 2002.
285 • NOVEMBER 2003 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on June 14, 2017
ments. Coronary microembolization with 300-␮m-diameter microspheres had a significantly higher impact
than that with 100-␮m-diameter microspheres because
300-␮m-diameter microspheres occlude more proximal
branches of the vasa vasorum tree structure, which
leaves all hierarchically more distal-located vasa vasorum segments nonperfused, or they even occlude major
LAD branches, which leaves complete vasa vasorum
trees nonperfused [as we have shown earlier (5), LAD
branches feed the majority of arterial vasa vasorum].
Our subanalysis with differentiation of venous and
arterial vasa vasorum (Fig. 5), by virtue of their connection to either an artery or vein, shows that especially the perfusion of venous vasa vasorum on the
myocardial side of the coronary vessel wall is altered
under physiological intraluminal pressure. The consistently lower vasa vasorum densities on the myocardial
side of the coronary arteries [in physiologically (antegradely) perfused LADs; Fig. 3] suggest a possible
relationship to plaque formation predominant on the
myocardial side of coronary arteries (7, 15, 19). Future
studies will need to be performed to determine whether
the distribution of low-vasa-vasorum-density territories precisely correlates with early, patchy, atherosclerotic changes in the coronary artery wall (11). Nonetheless, Nakata et al. (13) showed in rats that the
obstruction of the venous side of aortic vasa vasorum
causes atherosclerotic lesions. In conjunction with the
“response to retention” model, as recently proposed by
Skalen et al. (18), the results of Nakata et al.’s study
and the present study emphasize the need to further
explore the relative role of supply to, and drainage
from, the coronary vessel wall in disease.
Finally, experimental microembolization does not
only help us to define the functional structure of vasa
vasorum, it is, moreover, a model worth pursuing because it shows the immediate impact of disturbances
within the coronary artery’s microcirculation on the
perfusion of its vessel wall. Earlier experimental studies (1, 12, 13) on various animal arteries (not coronary
arteries) have shown that the impairment of nutrient
flow through the vasa vasorum leads to atherosclerotic
lesions. Moreover, as others (4, 20) have shown that
coronary microembolization leads to an increased myocardial inflammatory response, which is associated
with a progressive contractile dysfunction, it seems
plausible that microembolization of coronary vasa vasorum may also induce microinflammation in the coronary vessel wall. This possibility, however, has to be
tested in further chronic experiments before any stronger conclusions of the role of microembolization in
coronary atherosclerosis can be drawn. Nonetheless,
knowing the functional structure of vasa vasorum may
elucidate their role in determining the size and location of atherosclerotic lesions.
H2025
H2026
FUNCTIONAL STRUCTURE OF CORONARY VASA VASORUM
19. Strunk W, Burrig KF, and Hort W. Quantitative morphologic
studies for localizing arteriosclerotic plaques in the circumference of coronary arteries. Z Kardiol 79: 273–278, 1990.
20. Thielmann M, Dorge H, Martin C, Belosjorow S,
Schwanke U, van De Sand A, Konietzka I, Buchert A,
Kruger A, Schulz R, and Heusch G. Myocardial dysfunction
with coronary microembolization: signal transduction through a
sequence of nitric oxide, tumor necrosis factor-alpha, and sphingosine. Circ Res 90: 807–813, 2002.
21. Tsikaras DP, Natsis K, Hytiroglou P, Lazos L, and Gigis P.
Microanatomy of the vasa vasorum of the human thoracic aorta:
22.
23.
24.
25.
a study utilizing a polyester resin casting technique. Morphologie 81: 21–22, 1997.
VanNostrand RCY. VanNostrand’s Scientific Encyclopedia.
New York: VanNostrand, 1976.
Von Ludinghausen M, Ohmachi N, Besch S, and Mettenleiter A. Atrial veins of the human heart. Clin Anat 8: 169–189,
1995.
Williams JK and Heistad DD. The vasa vasorum of the arteries. J Mal Vasc 21: 266–269, 1996.
Zamir M and Silver MD. Vasculature in the walls of human
coronary arteries. Arch Pathol Lab Med 109: 659–662, 1985.
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on June 14, 2017
AJP-Heart Circ Physiol • VOL
285 • NOVEMBER 2003 •
www.ajpheart.org