1684 The Journal of Clinical Investigation http://www.jci.org Volume 113 Number 12 June 2004
Noninvasive imaging of myocardial
angiogenesis following experimental
David F. Meoli,1 Mehran M. Sadeghi,1,2 Svetlana Krassilnikova,1,2 Brian N. Bourke,1
Frank J. Giordano,1 Donald P. Dione,1 Haili Su,1 D. Scott Edwards,3 Shuang Liu,3
Thomas D. Harris,3 Joseph A. Madri,4 Barry L. Zaret,1,5 and Albert J. Sinusas1,5
1Division of Cardiovascular Medicine, Department of Internal Medicine, Yale University School of Medicine, New Haven, Connecticut, USA.
2Section of Cardiology, Department of Veterans Affairs Connecticut Healthcare Center, West Haven, Connecticut, USA. 3Bristol-Myers Squibb, North Billerica,
Massachusetts, USA. 4Department of Pathology and 5Department of Diagnostic Radiology, Yale University School of Medicine, New Haven, Connecticut, USA.
Noninvasive imaging strategies will be critical for defining the temporal characteristics of angiogenesis and
assessing efficacy of angiogenic therapies. The αvβ3 integrin is expressed in angiogenic vessels and represents
a potential novel target for imaging myocardial angiogenesis. We demonstrated the localization of an indium-
111–labeled (111In-labeled) αvβ3-targeted agent in the region of injury-induced angiogenesis in a chronic rat
model of infarction. The specificity of the targeted αvβ3-imaging agent for angiogenesis was established using
a nonspecific control agent. The potential of this radiolabeled αvβ3-targeted agent for in vivo imaging was then
confirmed in a canine model of postinfarction angiogenesis. Serial in vivo dual-isotope single-photon emis-
sion–computed tomographic (SPECT) imaging with the 111In-labeled αvβ3-targeted agent demonstrated focal
radiotracer uptake in hypoperfused regions where angiogenesis was stimulated. There was a fourfold increase
in myocardial radiotracer uptake in the infarct region associated with histological evidence of angiogenesis and
increased expression of the αvβ3 integrin. Thus, angiogenesis in the heart can be imaged noninvasively with an
111In-labeled αvβ3-targeted agent. The noninvasive evaluation of angiogenesis may have important implications
for risk stratification of patients following myocardial infarction. This approach may also have significant clini-
cal utility for noninvasively tracking therapeutic myocardial angiogenesis.
Angiogenesis represents the formation of new capillaries by cel-
lular outgrowth from existing microvessels (1) and occurs as
part of the natural healing process following ischemic injury.
Angiogenesis is associated with postinfarct remodeling and has
important implications for prognosis following myocardial infarc-
tion. Therefore, the noninvasive evaluation of angiogenesis may
help predict left ventricular remodeling and permit risk stratifi-
cation of patients following myocardial infarction. In addition,
newer therapies for treatment of ischemic heart disease involve
stimulation of this natural response. Many approaches to thera-
peutic myocardial angiogenesis have been explored, including
gene therapy, intramyocardial administration of angiogenic fac-
tors, and transmyocardial revascularization (2–10). While treat-
ment with factors such as VEGF and bFGF has produced encour-
aging results in animal models (2–4) and in early clinical trials
(7, 11), double-blind, placebo-controlled clinical studies to date
have resulted in less-favorable outcomes (8, 10). These seemingly
contradictory findings may be directly related to the limitations
of existing clinical methods directed at noninvasive evaluation of
the physiological effects of angiogenesis, which may be very focal
or nontransmural and therefore difficult to detect noninvasively.
Unfortunately, standard cardiac noninvasive imaging strategies
have not been effective in evaluation of cardiac angiogenesis (12).
The development of more sensitive and specific imaging strategies
for the direct evaluation of myocardial angiogenesis, and tracking
this process noninvasively, will be critical in evaluation of patients
following myocardial infarction and in assessing new angiogenic
therapy for the heart.
The angiogenic response is modulated by the composition of
the ECM and intercellular adhesions, including integrins (13,
14). Integrins are a family of heterodimeric cell surface receptors
capable of mediating an array of cellular processes, including cell
adhesion, migration, proliferation, differentiation, and survival
(15). During angiogenesis endothelial cells must adhere to one
another and to the ECM in order to construct and extend new
microvessels. The specific αvβ3 integrin has been identified as a
critical modulator of angiogenesis (14).
Angiogenic vessels demonstrate increased expression of the
αvβ3 integrin; therefore the αvβ3 integrin represents a potential
novel target for directly imaging angiogenesis (16, 17). Inves-
tigators have proposed the potential noninvasive detection of
tumor angiogenesis in vivo using magnetic resonance imaging
and a paramagnetic contrast agent targeted to endothelial αvβ3
via the mAb for the αvβ3 integrin (LM609) (16). Haubner et al.
reported the synthesis and favorable imaging characteristics
of a iodine-125–labeled (125I-labeled) αvβ3 antagonist (17–19).
Harris et al. recently reported the high affinity and selectivity
of an indium-111–labeled quinolone (111In-RP748) for the αvβ3
Nonstandard abbreviations used: indium-111 (111In); iodine-125 (125I); left anterior
descending (LAD); magnetic resonance (MR); 2-methoxy-2-methylpropyl-isonitrile
(sestamibi); positron emission tomography (PET); single-photon emission–computed
tomographic (SPECT); Technetium-99m (99mTc); thallium-201 (201Tl); 2,3,5-triphenyl-
2H-tetrazolium chloride (TTC).
Conflict of interest: D.S. Edwards, S. Liu, and T.D. Harris were employees of Bristol-
Myers Squibb at the time the studies were performed.
Citation for this article: J. Clin. Invest. 113:1684–1691 (2004).
The Journal of Clinical Investigation http://www.jci.org Volume 113 Number 12 June 2004
integrin using assays of integrin-mediated adhesion (20). These
investigators also demonstrated a rapid blood clearance and
favorable biodistribution of 111In-RP748 and the feasibility for
tumor imaging. We recently demonstrated that 111In-RP748 binds
preferentially to activated αvβ3 on endothelial cells in vitro and
exhibits favorable binding characteristics for in vivo imaging (21).
These preliminary in vitro studies and work in imaging tumor
angiogenesis supports the potential for radiolabeled targeting of
αvβ3 for imaging of myocardial angiogenesis.
We hypothesized that the angiogenic process in the heart can be
directly tracked noninvasively by single-photon emission–comput-
ed tomographic (SPECT) imaging of radiolabeled ligands targeted
at the αvβ3 integrin. The purpose of the current study was to eval-
uate 111In-RP748, a radiolabeled quinolone targeted at αvβ3, for
imaging of ischemia-induced angiogenesis following ischemic inju-
ry in established rodent and canine models of myocardial infarc-
tion. The chemical structure of 111In-RP748 is shown in Figure 1A.
We evaluate the uptake of 111In-RP748 following ischemic injury,
in relation to immunohistochemical markers of the angiogenic
process and myocardial perfusion. The specificity of myocardial
111In-RP748 uptake was demonstrated in a chronic rodent model
of injury-induced myocardial angiogenesis using paired control
experiments with an analogue compound (Figure 1B), which dem-
onstrated no in vitro specificity for the αvβ3 integrin. Subsequent
studies employing a chronic canine model of injury-induced myo-
cardial angiogenesis demonstrated the feasibility of SPECT imag-
ing for noninvasive detection of the αvβ3 integrin.
All experimental studies were performed with approval of the
Institutional Animal Care and Use Committee, according to
the guiding principles of the American Physiological Society on
research animal use.
Rodent surgical preparation. A chronic rat model of infarction was
employed using procedures modified from those previously
reported (22). Male Sprague-Dawley rats (200–250 g) were anes-
thetized with isoflurane, and the hearts were exposed by a limited
left anterolateral thoracotomy. A 6-0 proline ligature was used to
occlude the left anterior descending (LAD) coronary artery 7 mm
distal to its origin. Serial brief preconditioning occlusions were
performed to suppress ventricular dysrhythmias, followed by a 45-
minute experimental occlusion. The coronary occlusion was then
released, and the chest was closed in layers.
Rodent experimental protocol. Two weeks after myocardial infarc-
tion, rats (n = 11) were injected intravenously with either 111In-
RP748 (1.2 ± 0.1 mCi), an agent targeted at the αvβ3 integrin, or
111In-RP790 (1.2 ± 0.1 mCi), a nonspecific control agent. Ninety
minutes after injection of one of the 111In-labeled agents, rats were
injected with thallium-201 (201Tl; 0.78 ± 0.06 mCi) for evaluation
of relative myocardial perfusion. Rats were euthanized 30 minutes
after 201Tl injection for postmortem analysis of myocardial tissue
activity. The myocardial uptake of the 111In-labeled agents was
compared to relative 201Tl perfusion and immunohistochemical
markers of the angiogenic process.
Postmortem analysis of rat hearts. Rodent hearts were excised and
filled with dental molding material (alginate impression mate-
rial, type II – normal set; Quala Dental Products, Milford, Dela-
ware, USA) to facilitate cutting of the hearts into uniform 2-
mm-thick slices. Selected slices were frozen and sectioned for
immunohistochemistry of both lectin (marker of endothelium) αv
and β3. Unstained heart slices were then cut into eight transmural
segments for γ-well counting of 201Tl and 111In activity, using two
energy windows (201Tl: 60–90 keV; 111In: 170–300 keV). Raw counts
were corrected for spill-up/spill-down, background, decay, and
weight. The corrected counts were normalized to a nonischemic
region of the heart, and the segments were segregated by myocardi-
al 201Tl retention into one of four groups: 0–40%, 41–60%, 61–80%,
or 80–120% nonischemic.
Studies were also performed in dogs (n = 6) with injury-induced
myocardial angiogenesis in order to better define the expression
of the αvβ3 integrin in the region of myocardial injury, to better
localize the uptake of 111In-RP748 in relationship to myocardial
injury and perfusion, and to establish the feasibility of in vivo
imaging of myocardial angiogenesis with 111In-RP748.
Canine surgical preparation. Six fasted adult mongrel dogs
were injected with 10 mg/kg thiopental sodium intravenously
and were intubated. Animals were placed on a respirator and
mechanically ventilated with nitrous oxide and oxygen (3:1) and
0.5–1.5% halothane. Dogs underwent cardiac catheterization via
the right femoral artery, using a standard coronary guide cathe-
ter, and a baseline angiogram was obtained. Either the proximal
LAD or left circumflex coronary artery was occluded for 2 hours
using a balloon angioplasty catheter (diameter: 2.5 mm; length:
20 mm) under fluoroscopic guidance. The balloon was inflated
to a pressure of 4 atm. An angiogram was performed immedi-
ately after inflation and every 15 minutes thereafter to confirm
coronary occlusion. The coronary occlusion was released after 2
hours, and coronary reflow was confirmed angiographically. An
angiogram was performed prior to each nuclear-imaging session
to confirm vessel patency.
Structure of 111In-RP748, a quinolone targeted at αvβ3 integrin (A),
and control compound (B).
1686 The Journal of Clinical Investigation http://www.jci.org Volume 113 Number 12 June 2004
In vivo SPECT imaging of 111In-labeled αvβ3 and perfusion. All dogs
(n = 6) underwent dual isotope SPECT imaging with 111In-RP748
and Technetium-99m–labeled (99mTc-labeled) 2-methoxy-2-methol-
propyl-isonitrile (sestamibi). Dogs were imaged following 6–10
hours of reperfusion, as well as at 1 and 3 weeks after reperfusion.
Following intravenous injection of 111In-RP748 (6.6 ± 1.0 mCi),
six serial 15-minute SPECT images were acquired starting at 15
minutes after injection. The last SPECT acquisition was initiated
90 minutes after injection. Images were acquired using two energy
windows, 180 keV ± 7.5% and 252 keV ± 10%, grouped into a single
image. All nuclear imaging was done on a Dual-Head Gamma Cam-
era (GE Millennium; General Electric Corp., Waukesha, Wiscon-
sin, USA), using medium energy parallel-hole collimators. SPECT
images were acquired in continuous advance mode, 15 seconds
per 3° frame, with a zoom of 1.77 and 64 × 64 matrix. Follow-
ing the 111In-RP748 SPECT imaging, 22.3 ± 5.4 mCi of 99mTc-ses-
tamibi was injected. A 15-minute SPECT 99mTc-sestamibi image
was acquired 15 minutes after injection to provide a reference
perfusion image facilitating reconstruction of the 111In-RP748
“hot spot” image. Acquisition parameters were identical to those
used for the 111In-RP748 SPECT acquisitions, except for an energy
window center at 140 keV (± 7.5%).
All SPECT images were batch reconstructed using standard fil-
tered back projection (order 4, cutoff 0.28 × Nyquest frequency)
without attenuation or scatter correction. To quantify the rela-
tive signal-to-noise ratio in our reconstructed SPECT images,
equal-size regions of interest were drawn over the targeted hot
spot and remote myocardial regions. The remote myocardial
region was localized within the myocardium on the
targeted image based on a location defined by the
registered perfusion images.
Postmortem analysis in canine model. Three weeks
after reperfusion, following the last imaging session,
dogs were euthanized and hearts rapidly excised.
Transmural myocardial sections were cut from the
normal and central infarct regions of the heart, and
tissue from each biopsy was split into two transmural
sections. One half was fixed in 10% formalin solu-
tion and later embedded in paraffin. The other half
of each biopsy was frozen in embedding compound
(Tissue-Tek OCT; Sakura Finetek USA Inc., Torrance,
California, USA) using a bath of 2-methylbutane
cooled on dry ice.
Ex vivo imaging of 111In-labeled αvβ3 and perfusion.
After left ventricle tissue was obtained from biopsies,
the heart was cast with dental impression material
and sliced into uniform short axis slices (5 mm). Cast
slices were placed in direct contact with the medium
energy collimator surface of the gamma camera in
order to obtain registered sequential high-resolution
(256 × 256) 99mTc-sestamibi and 111In-RP748 images
using the previously defined energy windows.
Postmortem analysis of canine hearts. Heart slices were
stained with a buffered solution of 2,3,5-triphenyl-2H-
tetrazolium chloride (TTC) at 38°C to identify myo-
cardial infarction. Digital photographs were taken of
both the apical and basal surfaces of all stained slices.
Heart slices were stripped of epicardial fat and sur-
face vessels, cut into eight radial pies, and each pie was
divided transmurally into epicardial and endocardial
pieces for γ-well counting using two energy windows (99mTc:
130–150 keV; 111In: 170–300 keV). Raw γ counts were corrected
for background, decay, spill-up/spill-down and were expressed as
counts per minute per gram of tissue. Activity in each myocardial
sample was normalized to the activity in a nonischemic region of
the heart. Tissue activities were grouped according to normalized
99mTc-sestamibi activity into one of four groups: 0–40%, 41–60%,
61–80%, or 81–120% nonischemic.
Histological and immunohistochemical analysis
Formalin-fixed myocardial sections were stained with Mas-
son’s trichrome using standard techniques to define the
area of fibrosis. Cryostat sections (5 μm thick) were used for
immunohistochemical staining. For our immunohistochemical
analysis of rat myocardial angiogenesis, sections stained with a
biotinylated endothelial-specific lectin Bandeiraea Simplicifo-
lia Lectin I (Vector Laboratories, Burlingame, California, USA).
Since no specific Ab is available for rat αvβ3 integrin, we per-
formed separate staining for αv (rabbit anti-integrin αv subunit
polyclonal Ab; Chemicon International, Temecula, California,
USA) and β3 (mouse anti-rat integrin β3 mAb; BD Biosciences
PharMingen, San Diego, California, USA) integrins. Sections
were then incubated with respective biotinylated secondary Ab’s.
The signal was detected using HRP avidin (Elite ABC VectaStain
Kit; Vector Laboratories). The peroxidase label was developed
using 3-(2-aminoethyl)carbazole (Red AEC Kit; Vector Labora-
tories), and the sections were counterstained with hematoxylin.
Analysis of angiogenesis in dogs was accomplished with the
Immunohistochemical analysis in chronic rat model. Example of lectin, αv, and β3
staining is shown for normal and infarct region (A) in rats 2 weeks following infarction.
The infarct region demonstrated an increase in capillary density and arterioles, as well
as increased staining for αv and β3. This was confirmed by quantitative analysis (B).
*P < 0.05 vs. normal.
The Journal of Clinical Investigation http://www.jci.org Volume 113 Number 12 June 2004
biotinylated endothelial-specific lectin, Bandeiraea Simplicifolia
Lectin I (Vector Laboratories). For immunostaining of the αvβ3
integrin in dogs, we used a specific mouse mAb for the αvβ3
integrin (LM609; Chemicon International).
Sections were analyzed quantitatively for extent (percentage
of area) of positive staining using computer algorithms devel-
oped and validated in our laboratory, which take into account
the image hue, saturation, and intensity color map. Quantitative
values for infarct and control regions in each animal were calcu-
lated by averaging results for four representative high-powered
fields (×200) in each region.
Canine cell culture studies
Canine femoral vein endothelial cells were isolated by scraping
and were cultured as previously described (21). A cy3-labeled
homologue (TA145) of 111In-RP748 was used to confirm and
localize endothelial cell binding. The localization of TA145 was
compared with that of LM609. LM609 is a mAb for the αvβ3
integrin, which has been previously demonstrated to bind to
canine microvascular endothelial cells on frozen tissue sections
from reperfused infarcted regions (23). The localization of TA145
and LM609 was also compared with an isotype-matched control
Ab of LM609 (MOPC21).
All data are presented as mean plus or minus SD. Comparisons
between two groups were made using either a paired or unpaired
Student’s t test. Differences between groups were considered sig-
nificant at P values less than 0.05 (two-tailed).
Rodent chronic myocardial infarct experiments
Studies were performed in rats (n = 11) with injury-induced myo-
cardial angiogenesis. Immunohistochemical staining with an
endothelial-specific lectin confirmed angiogenesis in the infarct
region. In the absence of a specific anti-rat αvβ3 Ab, αv and β3
γ-Well counting of myocardial radiotracer activity in relationship to 201Tl
uptake in the chronic rat model. Data are shown for rats injected with
either 111In-RP748 (RP748) or control compound. Decreased myocar-
dial 201Tl uptake was consistently observed in the anterolateral wall, as
shown in representative myocardial count profiles (A). Cntrl, control;
ant, anterior; sept, septal; post, posterior; lat, lateral. Uptake of 111In-RP748
was highest in infarcted regions with reduced 201Tl retention. In contrast,
myocardial uptake of the control compound tracked 201Tl perfusion. On
average the relative myocardial retention of 111In-RP748 in the post-
ischemic and infarcted regions was nearly twice that in regions with
normal 201Tl perfusion; however, no selective retention of the nonspe-
cific control compound was observed (B). *P < 0.05 vs. 81–120%;
#P < 0.05 vs. control.
Immunohistochemical analysis in chronic canine model. Masson’s tri-
chrome staining demonstrated increased vascular density in the fibrotic
central infarct region, which was confirmed by lectin staining. The infarct
region demonstrated increased staining for αvβ3 using the LM609 Ab
(A). LM609 staining localized to the endothelial cells of capillaries
and endothelial cells, as well as smooth muscle cells of small arte-
rioles within the infarct region. Very little LM609 staining was in seen
in remote noninfarcted regions. These regional differences in staining
were confirmed by quantitative analysis (B). *P < 0.05 vs. normal.
1688 The Journal of Clinical Investigation http://www.jci.org Volume 113 Number 12 June 2004
staining in this same region were performed and demonstrated
increased expression of these integrins (Figure 2A). These changes
were confirmed by semiautomated quantitative analysis of the
immunohistochemical stains (Figure 2B). Following in vivo injec-
tion of 111In-RP748 and the perfusion tracer thallium-201 (201Tl),
γ-well counting of myocardial segments allowed quantification
of relative myocardial 111In-RP748 retention in relationship to
myocardial perfusion. The 111In-RP748 was selectively retained in
the infarcted regions with reduced 201Tl perfusion (Figure 3A). A
twofold increase in the retention of 111In-RP748 was observed in
most ischemic regions at 2 weeks after infarction (Figure 3B). In
contrast, a nonspecific control compound demonstrated no pref-
erential retention in the infarct region. The specific retention of
111In-RP748 in myocardial regions with injury-induced angiogenesis
suggests that 111In-RP748 might be a candidate radiotracer for the
noninvasive imaging of myocardial angiogenesis.
Canine chronic myocardial infarct experiments
Immunohistochemical staining of endothelium and αvβ3 integrin. Represen-
tative immunohistochemical stains are shown in Figure 4A. Lectin
histochemical staining for endothelium confirmed angiogenesis
in the central infarct region of the chronically infarcted dogs.
Immunohistochemical staining with LM609, a specific Ab for the
αvβ3 integrin, also confirmed increased expression of the receptor
in capillaries and of both the endothelium and smooth musculature
of arterioles within the infarct region. Increased capillary staining
was primarily seen in the peri-infarct region, while arteriolar smooth
muscle staining with LM609 was observed within the central infarct
region. Quantitative analysis demonstrated a significant increase in
vascular staining in the infarct region and increased activation of the
αvβ3 integrin in angiogenic vessels (Figure 4B). The relative increase
in LM609 staining in the ischemic region (ischemic/nonischemic
ratio) observed in the dogs significantly correlated (r = 0.84, P ≤ 0.05)
with the relative increase in myocardial uptake of 111In-RP748 within
these same ischemic regions.
Postmortem tissue analysis in canine model. γ-Well counting of canine
myocardial segments allowed quantification of the observed
increase in relative myocardial 111In-RP748 retention in relation-
ship to myocardial perfusion. Myocardial segments with decreased
perfusion, as determined by 99mTc-sestamibi retention, demon-
strate increased retention of 111In-RP748. Figure 5 illustrates the
circumferential changes in endocardial and epicardial 111In-RP748
activity in relationship to 99mTc-sestamibi activity and correspond-
ing ex vivo images in one of the dogs euthanized 3 weeks following
infarction. Tissue γ-well counting data are summarized in Figure
5B. Dogs with chronic infarction demonstrated almost a fourfold
increase in relative myocardial 111In-RP748 activity in the most
ischemic regions as defined by 99mTc-sestamibi.
Dual-isotope in vivo SPECT imaging of 111In-RP748 and 99mTc-ses-
tamibi perfusion. The heart rate and aortic pressure remained
stable in all chronically infarcted dogs that underwent serial
dual-isotope in vivo SPECT imaging. The average hemodynam-
ic values for these dogs are summarized in Table 1. Analysis of
the in vivo 111In-RP748 SPECT images required registration and
batch reconstruction of 111In-RP748 image data with 99mTc-ses-
tamibi perfusion images. All dogs demonstrated focal myocar-
dial retention of 111In-RP748 on in vivo SPECT imaging by 60
minutes after injection, at a time when blood pool activity had
cleared. The increase in myocardial 111In-RP748 activity corre-
lated with a 99mTc-sestamibi perfusion defect. The 111In-RP748
image quality was generally excellent, with a favorable heart-to-
background activity ratio. The targeted (hot spot) myocardial
activity-to-remote background activity ratio varied from 1:1 to
1.6:1. A representative series of 111In-RP748 and corresponding
99mTc-sestamibi SPECT images from a chronically infarcted
dog is shown in Figure 6. Serial dual-isotope 111In-RP748 and
99mTc-sestamibi SPECT imaging demonstrated retention of
111In-RP748 in the infarct region (see Figure 6A). A represen-
tative dual-isotope 111In-RP748 and 99mTc-sestamibi SPECT
image series is shown in Figure 6B. Based on our analysis of the
serial images, the maximum increase in myocardial 111In-RP748
retention appeared to occur at 1 week after infarction.
Well counting of myocardial radiotracer activity in relationship to
99mTc-sestamibi perfusion in chronic canine infarct model. Ex vivo
99mTc-sestamibi perfusion (left) and 111In-RP748 (right) images of
representative myocardial slices from an infarcted dog, 3 wks after
reperfusion, with the corresponding TTC-stained section (middle) (A).
Corresponding circumferential 99mTc-sestamibi (MIBI) and 111In-RP748
count profiles are shown for both endocardial (ENDO) and epicardial
(EPI) segments. Focal uptake of 111In-RP748 is seen in infarct region
on ex vivo images, and is confirmed by γ-well counting. Myocardial
segments from all dogs were segregated into four categories based
on relative 99mTc-sestamibi perfusion (percentage nonischemic).
Regional myocardial 111In-RP748 activity (percentage nonischemic)
was significantly increased 3 wks after reperfusion in the ischemic
regions of all dogs (B). #P < 0.05 vs. 81–120.
The Journal of Clinical Investigation http://www.jci.org Volume 113 Number 12 June 2004
Dual-isotope ex vivo imaging of 111In-RP748 and 99mTc-sestamibi
perfusion. The superior resolution of dual isotope ex vivo slice
images allowed better localization of myocardial 111In-RP748
retention within the central infarct area. Figure 6C illustrates the
focal uptake of 111In-RP748 within the nontransmural 99mTc-ses-
tamibi perfusion defect. Myocardial retention of 111In-RP748 was
apparent in the central infarct region. These images allowed direct
correlation of myocardial 111In-RP748 retention with postmortem
TTC staining for necrosis (see Figure 5A).
Canine cell culture studies. In cultured canine endothelial cells, a
cy3-labeled analogue of 111In-RP748 (TA145) localized to αvβ3
at focal cellular contacts in a distribution similar to LM609, an
established Ab for αvβ3 integrin (see Figure 7). This pattern of
localization was not seen with an isotype-matched control Ab.
These in vitro data support that our radiolabeled compound
does in fact bind to endothelial cells.
The 111In-labeled αvβ3-targeted agent (111In-RP748) may be a
valuable targeted marker for noninvasive imaging of angiogenesis
following nontransmural myocardial infarction. We employed
established chronic rat and canine models of myocardial infarc-
tion, which are known to produce nontransmural infarction and
peri-infarct ischemia resulting in myocardial angiogenesis. In
our rat model of chronic infarction, 111In-RP748 was selectively
retained in the infarcted regions with reduced 201Tl perfusion.
The specificity of 111In-RP748 retention in myocardial regions
with ischemia-induced angiogenesis was supported by the use of
In vivo and ex vivo 111In-RP748 and 99mTc-sestamibi (99mTc-MIBI) images from dogs with chronic infarction. Serial in vivo 111In-RP748 SPECT
short axis, vertical long axis (VLA), and horizontal long axis (HLA) images in a dog 3 wks after LAD infarction at 20 min and 75 min after injec-
tion in standard format (A). 111In-RP748 SPECT images were registered with 99mTc-MIBI perfusion images (third row). The 75-min 111In-RP748
SPECT images were colored red and fused with 99mTc-MIBI images (green) to better demonstrate localization of 111In-RP748 activity within
the heart (color fusion, bottom row). Right ventricular (RV) and left ventricular (LV) blood pool activity is seen at 20 min. White arrows indicate
region of increased 111In-RP748 uptake in anterior wall. This corresponds to the anteroapical 99mTc-sestamibi perfusion defect (yellow arrows).
Sequential 99mTc-sestamibi (top row) and 111In-RP748 in vivo SPECT HLA images at 90 min after injection (middle row) from a dog at 8 h (acute),
1 wk, and 3 wks after LAD infarction (B). Increased myocardial 111In-RP748 uptake is seen in the anteroapical wall at all three time points. Color
fusion 99mTc-MIBI (green) and 111In-RP748 (red) images (bottom row) demonstrate 111In-RP748 uptake within 99mTc-MIBI perfusion defect. Ex
vivo 99mTc-sestamibi (left) and 111In-RP748 (center) images of myocardial slices from a dog 3 wks after LAD occlusion, with color fusion image on
the right (C). Short axis slices are in the standard orientation. Yellow arrows indicate anterior location of nontransmural perfusion defect region;
white arrows indicate corresponding area of increased 111In-RP748 uptake.
1690 The Journal of Clinical Investigation http://www.jci.org Volume 113 Number 12 June 2004
a negative control compound, which demonstrated no increased
uptake in the infarct region. These initial rat studies suggested
that 111In-RP748 might be a candidate radiotracer for the non-
invasive imaging of myocardial angiogenesis. The use of a canine
model permitted the noninvasive serial evaluation of changes
in myocardial 111In-RP748 uptake in relationship to changes of
myocardial perfusion. The 111In-RP748 cleared rapidly from the
blood and demonstrated a favorable biodistribution for imaging
of myocardial angiogenesis. In vivo SPECT imaging demonstrat-
ed focal 111In-RP748 uptake in the infarct region associated with
increased activation of the αvβ3 integrin. Relative 111In-RP748
activity within the infarct was increased to 370% of nonischemic
myocardium at 3 weeks after reperfusion. Analysis of serial in vivo
SPECT images suggests that peak uptake of 111In-RP748 may in
fact occur closer to 1 week following reperfusion. Noninvasive
imaging strategies similar to that employed in this study may
be useful for evaluating angiogenesis and assessing efficacy of
angiogenic therapies in patients with chronic ischemia or follow-
ing myocardial infarction.
Role of αvβ3 integrin in myocardial angiogenesis
Integrins are αβ heterodimeric receptors that mediate divalent cat-
ion-dependent cell-cell and cell-matrix adhesion through tightly
regulated interactions with ligands (24). The integrins are involved
with many fundamental cellular processes, such as attachment,
migration, proliferation, differentiation, and survival. The αvβ3
integrin allows endothelial cells to interact with a wide variety
of ECM components. The mAb against αvβ3 (LM609) inhibits
angiogenesis by selectively promoting apoptosis of vascular cells
(13). Therefore, the αvβ3 integrin plays a critical role in signaling
events vital to survival of vascular cells undergoing angiogenesis.
Enhanced expression of the αvβ3 integrin has been previous-
ly observed in new blood vessels involved with angiogenesis
(14). Endothelial cell migration is one of the critical steps in
angiogenesis that specifically requires αvβ3 integrins (25).
We observed increased expression of αvβ3 integrin as late as 3
weeks following ischemic injury in endothelial and VSMCs with-
in the central infarct area using immunohistochemical analysis
with LM609. Brooks et al. suggested that peak expression of αvβ3
occurs 12–24 hours after initiation of angiogenesis with bFGF
(14). Sun et al. recently demonstrated increased expression of β3 in
rats following infarction. In these studies of permanent coronary
occlusion, β3 mRNA and protein peaked at 7 days after infarc-
tion and localized in the endothelial and smooth muscle cells
in peri-infarct vessels (26). Additional experimental studies will
be required to define the duration of αvβ3 integrin expression
and/or activation following ischemic injury and under condi-
tions of stimulated angiogenesis.
Imaging of αvβ3 integrin
Investigators have previously proposed noninvasive imaging
of tumor angiogenesis with compounds specifically targeted
at the αvβ3 integrin, using magnetic resonance (MR) imaging
(16), positron-emission tomography (PET) (19), and SPECT
imaging (17). No prior studies have attempted αvβ3 integrin-
targeted imaging of myocardial angiogenesis.
Targeted SPECT imaging of myocardial angiogenesis offers
many advantages over clinical indices or imaging with other
modalities. Rather than measuring an effect or consequence
of angiogenesis, targeted SPECT imaging directly evaluates
the angiogenic process. This is especially important considering
the strong placebo effect that has been suggested in clinical trials
of angiogenic factors. While clinical studies of angiogenic thera-
pies have demonstrated minor increases in SPECT perfusion, we
have demonstrated an approximate two- to fourfold increase in
radiotracer uptake in regions of ischemia-induced myocardial
angiogenesis. The dual-isotope SPECT perfusion and 111In-RP748-
imaging approach described in this study could easily be adapted
for clinical imaging using existing equipment. In this study, we
used an 111In-labeled quinolone targeted at αvβ3 for imaging of
ischemia-induced angiogenesis, which permitted serial in vivo
SPECT imaging in a canine model of myocardial infarction. Image
quality for clinical application could be further improved by use of
a 99mTc-labeled compound targeted at the αvβ3 integrin in combi-
nation with 201Tl perfusion. This radiotracer combination would
allow use of higher-resolution low-energy collimators. Absolute
quantification of myocardial 111In-RP748 uptake will ultimately
require correction for attenuation, scatter, and partial volume
errors. These corrections will become feasible clinically with the
availability of hybrid SPECT and x-ray computed tomography
imaging systems that will allow for correction of these errors.
Angiogenesis plays an important role in infarct healing and left
ventricular remodeling following myocardial infarction, and
infarct remodeling has important implications for the progno-
sis following myocardial infarction. Therefore, the noninvasive
evaluation of angiogenesis may have important implications for
prediction of left ventricular remodeling and risk stratification of
patients following myocardial infarction. The ability to stimulate
myocardial angiogenesis also has applications in the treatment of
a host of cardiac disorders involving myocardial ischemia, includ-
Evaluation of αvβ3 integrin localization in cultured canine endothelial
cells. LM609, an established mAb of αvβ3 integrin, is localized at cell
contact points (left). A cy3-labeled analogue of 111In-RP748 (TA145) also
localized to αvβ3 at focal cellular contact points in a distribution similar
to LM609 (middle). No focal cellular uptake was seen with the isotype-
matched negative control Ab (MOPC21, right).
Hemodynamics, chronic canine studies
Heart rate (bpm)
Systolic AoP (mmHg)
Diastolic AoP (mmHg)
102 ± 19
101 ± 24
71 ± 11
98 ± 19
114 ± 27
75 ± 20
92 ± 24
99 ± 21
72 ± 23
bpm, beats per minute; AoP, aortic pressure.
research article Download full-text
The Journal of Clinical Investigation http://www.jci.org Volume 113 Number 12 June 2004
ing chronic coronary artery disease and myocardial infarction.
Numerous approaches to exogenously stimulate vessel growth
in the myocardium have been explored (2, 5–10). While to date a
great deal of effort has been focused on the development of strate-
gies to stimulate angiogenesis in the heart, there remains the need
for an effective, clinically feasible method for the evaluation of
these therapeutic strategies. Most clinical studies have relied on
improvement of clinical indicators (i.e., increased treadmill exer-
cise times, reduction in anginal classification), or evaluation of
the physiological consequences of angiogenesis, as indirect evi-
dence of myocardial angiogenesis. SPECT, PET, or MR imaging
is frequently used to detect changes in myocardial perfusion or
function as an indirect physiological indicator of angiogenesis.
While these imaging approaches are very useful in identifying
regions of moderately to severely diminished perfusion, they
may lack the sensitivity to discriminate the potentially small
physiological changes that might be expected to result from
stimulated angiogenesis in the myocardium. In animal models,
immunohistochemical methods and vessel counting are the gold
standard for measuring angiogenesis. Immunohistological stain-
ing of the endothelium is very effective in identifying blood ves-
sels in a sample of myocardium. These methods involve harvest-
ing of tissue samples and, of course, would be inappropriate for
clinical use. Targeted noninvasive SPECT hot spot imaging of the
angiogenic process, however, similar to that described here, may
be more sensitive for assessing efficacy of angiogenic therapies in
patients with chronic ischemia or following myocardial infarction
and permit characterization of angiogenesis in humans.
Targeted SPECT hot spot imaging of myocardial angiogenesis
with 111In-RP748 may provide increased sensitivity for noninva-
sive evaluation of angiogenesis over methods that evaluate the
physiological consequences of angiogenesis. In experimental
studies of angiogenesis, targeted SPECT imaging of the αvβ3
integrin offers the ability to evaluate serially the angiogenic pro-
cess over time in both small and large animal models. In our
experimental models of ischemia-induced angiogenesis, we have
also confirmed the increased expression of the αvβ3 integrin with
immunohistochemistry. While immunohistochemical techniques
permit evaluation of angiogenesis, information on the time course
of angiogenic processes can be obtained only by sacrificing many
animals at various time points. Although we can not exclude the
possibility of additional binding of 111In-RP748 to inflammatory
cells within the infarct region, we have demonstrated clear specific
binding of a cy3-labeled homologue of 111In-RP748 to focal cel-
lular contact points in cultured endothelial cells.
In this study we have demonstrated the ability of 111In-
RP748 SPECT imaging for serial evaluation of injury-induced
angiogenesis. Importantly, the imaging approaches developed for
these models of chronic infarction may directly translate to evalu-
ation of angiogenesis in patients.
This research was supported by a grant-in-aid from the American
Heart Association (0050516N) and a grant from the NIH, Nation-
al Heart, Lung, and Blood Institute (1R01 HL-65662), both to A.J.
Sinusas. The radiotracer was provided by Bristol-Myers Squibb.
We gratefully acknowledge the technical assistance of Xiao-Yu Hu,
Ion Jovin, Jason Soares, Christopher Weyman, and Anjali Nath.
Received for publication October 21, 2003, and accepted in revised
form April 20, 2004.
Address correspondence to: Albert J. Sinusas, Yale University
School of Medicine, Nuclear Cardiology, 3FMP, PO Box 208017,
New Haven, Connecticut 06520-8017, USA. Phone: (203) 785-
4915; Fax: (203) 737-1026; E-mail: Albert.Sinusas@yale.edu.
1. Battegay, E.J. 1995. Angiogenesis: mechanistic
insights, neovascular diseases, and therapeutic
prospects. J. Mol. Med. 73:333–346.
2. Giordano, F.J., et al. 1996. Intracoronary gene
transfer of fibroblast growth factor-5 increases
blood flow and contractile function in an ischemic
region of the heart. Nat. Med. 2:534–539.
3. Lazarous, D.F., et al. 1995. Effects of chronic sys-
temic administration of basic fibroblast growth
factor on collateral development in the canine
heart. Circulation. 91:145–153.
4. Unger, E., et al. 1994. Basic fibroblast growth fac-
tor enhances myocardial collateral flow in a canine
model. Am. J. Physiol. 266:H1588–H1595.
5. Schofield, P.M., et al. 1999. Transmyocardial laser
revascularisation in patients with refractory angi-
na: a randomised controlled trial. [erratum 1999,
353:1714]. Lancet 353:519–524.
6. Allen, K.B., et al. 1999. Comparison of
transmyocardial revascularization with medical
therapy in patients with refractory angina. N. Engl.
J. Med. 341:1029–1036.
7. Rosengart, T.K., et al. 1999. Angiogenesis gene ther-
apy: phase I assessment of direct intramyocardial
administration of an adenovirus vector expressing
VEGF121 cDNA to individuals with clinically sig-
nificant severe coronary artery disease. Circulation.
8. Hendel, R.C., et al. 2000. Effect of intracoronary
recombinant human vascular endothelial growth
factor on myocardial perfusion: evidence for a
dose-dependent effect. Circulation. 101:118–121.
9. Vale, P.R., et al. 2000. Left ventricular elec-
tromechanical mapping to assess efficacy of
phVEGF(165) gene transfer for therapeutic
angiogenesis in chronic myocardial ischemia.
10. Udelson, J.E., et al. 2000. Therapeutic angiogenesis
with recombinant fibroblast growth factor-2
improves stress and rest myocardial perfusion
abnormalities in patients with severe symptom-
atic chronic coronary artery disease. Circulation.
11. Losordo, D.W., et al. 1998. Gene therapy for myo-
cardial angiogenesis: initial clinical results with
direct myocardial injection of phVEGF165 as
sole therapy for myocardial ischemia. Circulation.
12. Simons, M., et al. 2000. Clinical trials in coronary
angiogenesis: issues, problems, consensus: An expert
panel summary. Circulation. 102:E73–E86. http://
13. Brooks, P., et al. 1994. Integrin avb3 antagonists
promote tumor regression by inducing apoptosis
of angiogenic blood vessels. Cell. 79:1157–1164.
14. Brooks, P., Clark, R., and Cheresh, D. 1994. Require-
ments of vascular integrin avb3 for angiogenesis.
15. Schwartz, M.A., Schaller, M.D., and Ginsberg, M.H.
1995. Integrins: emerging paradigms of signal
transduction. Annu. Rev. Cell Dev. Biol. 11:549–599.
16. Sipkins, D.A., et al. 1998. Detection of tumor
angiogenesis in vivo by alphaVbeta3-targeted mag-
netic resonance imaging. Nat. Med. 4:623–626.
17. Haubner, R., et al. 1999. Radiolabeled alpha(v)beta3
integrin antagonists: a new class of tracers for
tumor targeting. J. Nucl. Med. 40:1061–1071.
18. Haubner, R., et al. 2001. Glycosylated RGD-con-
taining peptides: tracer for tumor targeting and
angiogenesis imaging with improved biokinetics.
J. Nucl. Med. 42:326–336.
19. Haubner, R., et al. 2001. Noninvasive imaging
of alpha(v)beta3 integrin expression using 18F-
labeled RGD-containing glycopeptide and positron
emission tomography. Cancer Res. 61:1781–1785.
20. Harris, T.D., et al. 2003. Design, synthesis, and
evaluation of radiolabeled integrin v 3 receptor
antagonists for tumor imaging and radiotherapy.
Cancer Biother. Radiopharm. 18:631–645.
21. Sadeghi, M., et al. 2004. Detection of injury-
induced vascular remodeling by targeting acti-
vated alpha V beta 3 integrin in vivo. Circulation.
22. McNulty, P., and Luba, M. 1995. Transient
ischemia induces regional myocardial glycogen
synthase activation and glycogen synthesis in vivo.
Am. J. Physiol. 268:H364–H370.
23. Frangogiannis, N., et al. 2001. Induction and
suppression of interferon-inducible (IP)-10 in
reperfused myocardial infarcts may regulate
angiogenesis. FASEB J. 15:1428–1430.
24. Xiong, J.P., et al. 2001. Crystal structure of the
extracellular segment of integrin alpha Vbeta3.
25. Clyman, R. 1992. Beta 1 and beta 3 integrins have
different roles in the adhesion and migration
of vascular smooth muscle cells on extracellular
matrix. Exp. Cell Res. 200:272–284.
26. Sun, M., et al. 2003. Temporal response and
localization of integrins beta1 and beta3 in the
heart after myocardial infarction: regulation by
cytokines. Circulation. 107:1046–1052.