DOI:10.2214/AJR.06.0280
AJR 2007; 188:11-23
© American Roentgen Ray Society
Imaging of Angiogenesis: Clinical Techniques and Novel Imaging Methods
James M. Provenzale1
1 Department of Radiology, Duke University Medical Center, Box 3808, Durham, NC
27710.
Received March 14, 2006;
accepted after revision June 7, 2006.
Address correspondence to J. M. Provenzale.
CME
This article is available for CME credit. See
www.arrs.org
for more information.
Abstract
OBJECTIVE. A wide variety of antiangiogenic agents have been
developed for the treatment of neoplasms. Imaging studies play an important
role in assessing the effects of these treatments.
CONCLUSION. This review article introduces radiologists to features
of these therapies and the most important clinical and preclinical imaging
techniques for evaluating antiangiogenic agents.
Keywords: angiogenesis • dynamic MRI • molecular imaging • MR contrast agents • oncology • perfusion CT • perfusion-weighted MRI
Introduction
In the past few decades, the role of new blood vessel development
(angiogenesis) in the maintenance of health and development of various
pathologic states has become evident. For instance, angiogenesis has been
shown to play an important role in the development of collateral circulation
in the setting of atherosclerosis
[1]. In addition, an important
role for angiogenesis in the growth and metastasis of tumors has been well
established [2]. As a result, a
wide range of novel therapies directed against angiogenesis as a form of tumor
therapy have been developed. Although the number of antiangiogenesis trials in
humans is presently relatively small, it is growing at a substantial rate.
Because of the increasing importance of this topic, it is worthwhile for
radiologists to become familiar with types of antiangiogenic agents and the
methods for imaging results of these therapies.
Classification of Antiangiogenic Therapies
One categorization of novel antineoplastic drugs considers agents as being
targeted either against tumor cells and their microenvironment (e.g., cell
receptors, cell signaling systems, and oncogenes) or against tumor vascularity
[3]. The latter can be targeted
against existing tumor blood vessels (so-called vascular disrupting agents) or
against tumor blood vessel development (antiangiogenic agents). Many different
types of antiangiogenic agents exist. The reader is directed to a
government-sponsored Web site that provides a current tabulation of
antiangiogenic agents for detailed information
[4]. The site also provides the
following useful categorization of antiangiogenic agents: blockers of matrix
breakdown, direct inhibitors of endothelial cells, blockers of activation of
angiogenesis, inhibitors of endothelial-specific integrin and survival
signaling, and drugs with nonspecific mechanisms of action.
In parallel with the development of this wide array of therapeutic
strategies, a need has arisen for development of biomarkers that can be used
to monitor therapeutic effects. Because many antiangiogenic agents are not
cytotoxic but instead produce disease stabilization, measurement of tumor size
alone may be noninformative regarding therapeutic effect
[5]. For this reason, attention
has centered on the use of physiologic, rather than solely anatomic, imaging
techniques. However, because tumor necrosis and eventual diminution in size
may eventually occur even with tumor-stabilizing antiangiogenic agents, tumor
size is also considered in the evaluation of antiangiogenic agents.
Types of Biomarkers in Antiangiogenic Therapy
The biomarkers used to image the effects of therapeutic agents of any type
(i.e., not solely antiangiogenic agents) can be classified as either direct or
surrogate (indirect) in nature (see table at National Cancer Institute [NCI]
Web site [4]). A direct
biomarker is the actual target at which an agent is expected to exert its
effects. In the instance of antiangiogenic agents, the direct biomarker is the
actual blood vessels in the tumor, which are generally difficult to depict and
quantify using cross-sectional imaging techniques. For this reason, surrogate,
or indirect, markers are commonly used. Almost all direct markers are
available solely in animal models of tumors, whereas surrogate markers are
typically used in human studies. A surrogate end point has been described in
the following manner: "Investigators use surrogate endpoints when the
end point of interest is too difficult and/or expensive to measure routinely
and when they can define some other, more readily measurable end point, which
is sufficiently well correlated with the first to justify its use as a
substitute" [6]. Using
antiangiogenesis as an example, surrogate end points, or markers, consist of
measures of effects of increases or decreases in the numbers or types of blood
vessels—for example, the degree of contrast enhancement, assessments of
oxygen saturation, and MRI or CT measurements of blood volume.
Correspondence Between Imaging Agents and Site of Action of Antiangiogenic Therapies
In addition to the concepts of direct and surrogate end points, another
useful categorization of imaging biomarkers in antiangiogenic therapy has
recently been proposed [7].
This scheme is based on the proximity of the biomarker to the interaction of
the agent and molecular target. According to this classification, biomarkers
can reflect an interaction at the site of target inhibition, a global effect
on tumor microvasculature, or an effect on tumor metabolism (e.g., rate of
proliferation or cell death) (Table
1). Thus, the three markers range from those reflecting action at
the intended target to those that are distant, or downstream, from the
target.
Angiogenesis imaging techniques depicting interactions at the site of
target inhibition include direct imaging of blood vessels or of proteins and
receptors involved in the growth of blood vessels. Such techniques have
essentially entirely been used in animal models to date. An example of an
imaging agent reflecting such an interaction is one that becomes apparent when
a molecular interaction occurs (e.g., a "smart" probe that becomes
detectable only on binding with an angiogenesis growth factor membrane
receptor, as discussed in the following text).
Almost all techniques used for measuring the global effect on tumor
microvasculature and the effect on tumor metabolism (i.e., the second form of
biomarkers) in angiogenesis are indirect forms of imaging using surrogate
markers. Thus, these techniques do not image microvasculature itself but
instead image surrogate end points for microvasculature. For instance, many of
the commonly used perfusion imaging techniques for measuring global effects on
tumor microvasculature, such as dynamic contrast-enhanced perfusion MRI, do
not image vasculature (i.e., depict vessels) but instead allow inferences
about vascular characteristics. As an example, a decrease in cerebral blood
volume would be expected to correlate with a decrease in capillary
density.
MR spectroscopy and diffusion MRI are examples of techniques that reflect
the rate of tumor proliferation or cell death. These imaging techniques are
also indirect forms of imaging using surrogate markers, but they reflect
processes that are even further downstream from potential sites of target
inhibition than those measuring the global effect on tumor microvasculature.
As such, these techniques are also less specific in the information they
provide with regard to angiogenesis.
Categories of Contrast Agents
Similar to the classification of biomarkers into the three types outlined
is the distinction between various broad categories of contrast agents:
nonspecific contrast agents, targeted contrast agents, and smart contrast
agents [8]
(Table 1). Standard contrast
agents that are presently in clinical use in humans are examples of
nonspecific contrast agents. In general, these agents can provide information
solely about surrogate markers of tumor physiology (e.g., blood volume) and
thus have a relative lack of specificity. Targeted contrast agents, on the
other hand, are designed to bind to specific markers such as cell surface
proteins. This goal can be accomplished by means of linking the agent to
affinity ligands such as antibodies or peptides, which can provide relatively
high degrees of specificity. Smart probes have high specificity for their
target and undergo change on interacting with their target—for example,
the probes can alter signal characteristics after a physiologic event, such as
the tumor undergoing enzymatic cleavage. Like targeted contrast agents, smart
probes have a high affinity for their target. However, unlike targeted
contrast agents, smart probes undergo a change on binding with the target.
Targeted contrast agents and smart contrast agents can be used to image any
of the three types of biomarkers outlined previously but are the only types of
contrast agents that can be used for direct imaging of blood vessels.
Indirect Imaging of Angiogenesis
Indirect imaging of angiogenesis is more commonly performed than direct
techniques and is the type of angiogenesis imaging with which readers are most
likely to be familiar. For this reason, a discussion of the techniques used
for indirect imaging will precede the discussion of direct imaging.
Perfusion MRI
One of the major applications of perfusion MRI has been to assess
hemodynamic parameters of tumors. In turn, MRI is the technique that has
gained the most acceptance in terms of assessment of angiogenesis in humans.
The reader is directed to a number of excellent review articles that explain
the imaging principles underlying various techniques used for MR hemodynamic
assessment of tumors [3,
9-12].
Basically, hemodynamic imaging can be performed using either T1-weighted
techniques (generally termed "dynamic contrast-enhanced imaging")
or T2*-weighted techniques (generally termed "dynamic
susceptibility contrast imaging"). T1-weighted techniques can provide a
number of measures related to the rate of leakage of contrast
material—that is, permeability—and T2*-weighted
techniques are predominantly used to measure relative cerebral blood volume
(rCBV), which correlates with capillary density
[13].
View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —62-year-old man with glioblastoma multiforme. Depiction of
changes in leakage of contrast material (a reflection of degree of
permeability across blood-brain barrier) after administration of
antiangiogenic agent directed against vascular endothelial growth factor
(VEGF) receptor tyrosine kinase. Series of images shows additional physiologic
information provided by dynamic imaging relative to solely anatomic
information provided by conventional MRI. Because imaging findings solely
reflect effect of angiogenesis (i.e., permeability), this technique is an
example of indirect imaging of angiogenesis. Contrast-enhanced axial
T1-weighted image obtained before therapy shows large enhancing mass located
in left temporal lobe.
|
|
View larger version (103K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —62-year-old man with glioblastoma multiforme. Depiction of
changes in leakage of contrast material (a reflection of degree of
permeability across blood-brain barrier) after administration of
antiangiogenic agent directed against vascular endothelial growth factor
(VEGF) receptor tyrosine kinase. Series of images shows additional physiologic
information provided by dynamic imaging relative to solely anatomic
information provided by conventional MRI. Because imaging findings solely
reflect effect of angiogenesis (i.e., permeability), this technique is an
example of indirect imaging of angiogenesis. Color-coded map derived from
dynamic contrast-enhanced imaging sequence before therapy (corresponding to
A) shows degrees of contrast enhancement (changes in signal intensity)
normalized against signal intensity in normal tissue. Color-coded map
indicates relative degrees of permeability in each pixel, with blue pixels
representing signal intensity that is 120-149% of normal tissue, green pixels
representing 140-159% of normal tissue, and red pixels representing > 160%
of normal tissue. Note that most pixels are in > 160% range, indicating
marked leakage of contrast material.
|
|
The medical literature exploring the use of hemodynamic MRI for grading
brain tumors or comparing hemodynamic MRI measurements with findings obtained
using other advanced imaging techniques is substantial. In addition, a number
of articles have been written on the relative merits of individual MR
hemodynamic imaging paradigms, to which the reader is referred for details
[10,
11].
However, studies that specifically address the role of MRI for assessing
angiogenesis by comparing hemodynamic parameters against levels of
angiogenesis-promoting factors or measuring responses to antiangiogenic
therapies have largely been performed in animals
[14,
15]. These studies have shown
a promising role for MRI in this setting, although in many instances the
results have depended on the type of contrast material used. In one study,
investigators used microvascular permeability as a surrogate end point for
evaluating an antibody directed against the vascular endothelial growth factor
(VEGF) against breast cancer cells in nude mice and found that this therapy
produced both decreases in tumor growth rates and decreases in permeability
[14]. In another study in
which a monoclonal antibody to VEGF was administered to mice harboring an
implanted human cancer line, tumors grew more slowly in treated animals, but
the differences were not statistically significant. The investigators found a
significant decrease in MRI measurements of transendothelial permeability but
not fractional blood volume using a novel protein-binding contrast agent
[15]. However, no significant
differences were seen in either transendothelial permeability or fractional
blood volume when imaging was performed using either conventional MR contrast
material or albumin-gadopentetate dimeglumine.
View larger version (150K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C —62-year-old man with glioblastoma multiforme. Depiction of
changes in leakage of contrast material (a reflection of degree of
permeability across blood-brain barrier) after administration of
antiangiogenic agent directed against vascular endothelial growth factor
(VEGF) receptor tyrosine kinase. Series of images shows additional physiologic
information provided by dynamic imaging relative to solely anatomic
information provided by conventional MRI. Because imaging findings solely
reflect effect of angiogenesis (i.e., permeability), this technique is an
example of indirect imaging of angiogenesis. Contrast-enhanced axial
T1-weighted image obtained 30 days after initiation of therapy shows mild
decrease in size of mass.
|
|
View larger version (102K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D —62-year-old man with glioblastoma multiforme. Depiction of
changes in leakage of contrast material (a reflection of degree of
permeability across blood-brain barrier) after administration of
antiangiogenic agent directed against vascular endothelial growth factor
(VEGF) receptor tyrosine kinase. Series of images shows additional physiologic
information provided by dynamic imaging relative to solely anatomic
information provided by conventional MRI. Because imaging findings solely
reflect effect of angiogenesis (i.e., permeability), this technique is an
example of indirect imaging of angiogenesis. Color-coded map derived from
dynamic contrast-enhanced imaging sequence obtained 30 days after initiation
of therapy (corresponding to C) shows moderate decrease in total number
of color-coded pixels but, importantly, marked decrease in number of pixels at
middle and high end of contrast leakage range (green and red
pixels). Changes in degree of leakage of contrast material (reflecting changes
in degree of permeability) can be substantially greater than changes in
enhancing tumor size.
|
|
Experience with antiangiogenic agents in human tumors has been gained
primarily in the past few years
[16-20].
Initial studies used conventional MRI to measure tumor response as defined by
bidirectional diameter [18,
19]. More recently, perfusion
MRI measurements have come into use in this regard. In one study of human
liver metastases from colorectal carcinoma, researchers found a dose-related
decrease in permeability measurements after the administration of an inhibitor
of the VEGF receptor tyrosine kinase
[20]. Using the same agent in
one of the few trials of an antiangiogenic therapy for human brain tumors,
investigators showed dose-related early rCBV decreases and decreases in
contrast enhancement (thought to reflect permeability) on dynamic
contrast-enhanced MRI (Provenzale JM et al., presented at the 2002 meeting of
the Radiological Society of North America) (Figs.
1A,
1B,
1C, and
1D). In another study of
anti-VEGF antibody in patients with solid abdominal tumors, investigators
found decreases in the vascular permeability-surface area product at day 2
after beginning treatment that were not fully sustained at day 35
[21].
At this time, individual investigators differ on the relative merits of
measuring permeability rather than rCBV for assessing angiogenesis. Because
rCBV is generally understood to correlate with microvascular density, many
investigators consider this parameter to be a robust means of measuring
changes in degree of angiogenesis
[22]. However, one of the
important angiogenesis factors—that is, VEGF—is a potent vascular
permeability factor [23].
Therefore, it stands to reason that measuring rates of leakage of contrast
material across the blood-brain barrier might also be a valuable means of
measuring effects of antiangiogenic agents
[24] (Figs.
1A,
1B,
1C, and
1D). In fact, a rodent model of
brain tumors has shown decreases in permeability after treatment with a
monoclonal antibody directed against VEGF
[25]. The principles of that
study have subsequently been validated in a trial of anti-VEGF monoclonal
antibody therapy for rectal carcinoma in humans, in which changes in tumoral
interstitial fluid pressure served as a surrogate marker for permeability
[26].
Despite the theoretic possibilities just discussed, a number of
shortcomings have been noted using standard MR contrast agents for measuring
angiogenesis. One major shortcoming is that, using contrast-enhanced
T1-weighted imaging techniques, some degree of overlap of transendothelial
permeability values can be seen between benign and malignant non-CNS tumors
[27,
28]. Higher-molecular-weight
contrast agents and ultrasmall iron oxide particles also theoretically offer
some advantages relative to standard contrast agents and may provide a means
to circumvent this limitation
[29].
View larger version (139K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —Perfusion CT image of rectal carcinoma performed to assess
response to chemotherapy and radiation therapy. (Reprinted with permission
from Sahani DV, Kalva SP, Hamberg LM, et al. Assessing tumor perfusion and
treatment response in rectal cancer with multisection CT: initial
observations. Radiology 2005; 234:785-792
[36].) Color-coded blood flow
map in which areas of highest blood flow are shown in red, followed by yellow,
indicates regions of highest blood flow are scattered throughout much of
central portion of mass.
|
|
View larger version (143K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —Perfusion CT image of rectal carcinoma performed to assess
response to chemotherapy and radiation therapy. (Reprinted with permission
from Sahani DV, Kalva SP, Hamberg LM, et al. Assessing tumor perfusion and
treatment response in rectal cancer with multisection CT: initial
observations. Radiology 2005; 234:785-792
[36].) Color-coded
permeability-surface area product map—in which areas of highest
permeability-surface area product are shown in red and yellow—indicates
that regions of highest permeability are scattered throughout much of
tumor.
|
|
Perfusion CT
Animal studies have validated perfusion CT as a test for evaluating tumor
perfusion and permeability. In one study, investigators found a good
correlation between CT perfusion measurements and quantitative fluorescent
microsphere perfusion in tumor-bearing rats
[30]. In the same study, CT
permeability measurements were shown to correlate well with semiquantitative
Evans blue dye permeability estimates
[30]. In a rabbit tumor model,
dynamic CT measurements of cerebral blood flow under various physiologic
conditions were found to correlate well with results from ex vivo
microsphere-derived studies
[31].
The principles of analysis of CT-based perfusion techniques have been
extensively reviewed elsewhere
[31]. In brief, the most
commonly used analysis technique uses a deconvolution-based assessment in
which operator-derived regions of interest are placed on a representative
artery and a representative vein on the perfusion image to allow measurement
of input functions [31].
Comparison of the arterial and tissue time-attenuation curves, after
correcting for volume averaging, then allows computation of cerebral blood
flow. Alternatively, rCBV can be derived by comparing the areas under the
tissue and venous time-attenuation curves.
Relatively few studies of the use of CT for hemodynamic assessment of
tumors have been performed in humans, reflecting the general preference of
conventional MRI for clinical assessment of brain tumors. However, CT remains
the primary means for evaluating head and neck tumors for many physicians, and
perfusion CT might have particular relevance in patients with head and neck
tumors. In one study intended to determine whether tumors could be
distinguished from normal tissue on the basis of hemodynamic characteristics,
investigators found that untreated head and neck squamous cell cancers showed
significantly higher tissue blood volume, blood flow, and capillary
permeability-surface area product than those in normal tissue, and reduced
mean transit time from normal tissue
[32].
Because CT is the preferred method for imaging tumors of the abdomen and
thorax in many institutions, perfusion CT appears to have been used more
frequently than perfusion MRI for evaluating the response of abdominal tumors
to therapy. As an example, in one study investigators used perfusion CT to
characterize hepatic neoplasms and to assess the response of a small subset of
the tumors to transcatheter arterial chemoembolization
[33]. The authors were able to
distinguish contributions from portal perfusion from those due to arterial
perfusion and to monitor decreases in portal perfusion in some cases. For a
detailed review of perfusion CT techniques in assessing tumors, the reader is
referred to an excellent review article
[34].
In an important study, investigators implanted murine mammary cancer cells
in rodent liver and treated the animals with an antiangiogenic agent
consisting of a tyrosine kinase inhibitor of VEGF (SU5416) and then subjected
the animals to perfusion CT
[5]. The investigators found
that the mean tumor volume and the number of metastases were lower (although
not in a statistically significant manner) in treated animals than in control
animals. Although tumor microvascular density was significantly lower in
treated animals, vessel perimeter and vessel area were significantly higher,
as were blood flow, blood volume, and permeability-surface area product. The
results were interpreted as indicating that CT perfusion measurements reflect
changes in mature tumor vasculature (as opposed to the immature vessels
typically seen in regions of angiogenesis) and do not reflect changes in
microvascular density. This statement is based on the fact that perfusion CT
can detect only vessels that have blood flow. Although endothelial cells and
immature vessels contribute heavily to the microvascular density in regions of
a tumor that are particularly angiogenic, they will not be detectable on
hemodynamic imaging if they lack substantial blood flow.
View larger version (87K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —Use of photoacoustic imaging to depict angiogenesis after
subcutaneous injection of pancreatic tumor cells in a rat. Using this
technique, light is converted into sound by optical absorption. A laser serves
as light source; light, which is predominantly absorbed by hemoglobin, causes
small focal temperature increase and restriction in dilation of RBCs. In turn,
local pressure increases are generated, producing ultrasonic waves that allow
3D reconstruction of position of blood vessels. Regions of maximal acoustic
source strength are shown in red, followed by yellow. (Reprinted with
permission from Siphanto RI, Thumma KK, Kolkman RGM, et al. Serial noninvasive
photoacoustic imaging of neovascularization in tumor angiogenesis. Optics
Express 2005; 13:89-95
[45].) On day 3 after tumor
implantation, regions of neoangiogenesis are seen in green in one region of
image.
|
|
View larger version (124K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —Use of photoacoustic imaging to depict angiogenesis after
subcutaneous injection of pancreatic tumor cells in a rat. Using this
technique, light is converted into sound by optical absorption. A laser serves
as light source; light, which is predominantly absorbed by hemoglobin, causes
small focal temperature increase and restriction in dilation of RBCs. In turn,
local pressure increases are generated, producing ultrasonic waves that allow
3D reconstruction of position of blood vessels. Regions of maximal acoustic
source strength are shown in red, followed by yellow. (Reprinted with
permission from Siphanto RI, Thumma KK, Kolkman RGM, et al. Serial noninvasive
photoacoustic imaging of neovascularization in tumor angiogenesis. Optics
Express 2005; 13:89-95
[45].) On day 7 after tumor
implantation, regions of prominent angiogenesis are present throughout much of
image.
|
|
Despite the advances in perfusion CT listed here, studies using perfusion
CT to monitor tumors in humans after therapy are few
[35,
36]. One study designed to
predict the response of rectal cancer to chemotherapy and radiation therapy on
the basis of hemodynamic factors measured blood flow, mean transit time, and
vascular permeability-surface area product in tumors before and after therapy
[36] (Figs.
2A and
2B). Hemodynamic parameters
were compared with histopathologic findings indicating tumor regression at
surgery. Tumors with initial high blood flow and a short mean transit time
tended to respond poorly to chemotherapy and radiation therapy compared with
those exhibiting less profound hemodynamic alterations.
PET
The principal method for the indirect measurement of angiogenesis using PET
is evaluation of perfusion indexes using 15O water, which allows
assessment of cerebral blood flow. The perfusion data derived from
15O water measurements in tumors are reportedly similar to those
obtained using MRI and CT perfusion methods
[37,
38]. However, little
experience to date exists in the use of this technique for assessing untreated
tumors [39] or for assessing
tumor response to therapy
[37]. In one study of an
antiangiogenic agent in which liver metastases were imaged, significant
reductions in 15O water perfusion and 15O carbon
monoxide blood volume measurements were seen 30 minutes after drug
administration, although these changes were not generally sustained at 24
hours [40]. Direct
measurements of angiogenesis using PET methods in animal tumors have shown
some promise and are outlined later in this article.
Sonography
Sonography is an imaging technique that has shown great success for the
depiction of vascular architecture, patency, and flow rates in various organ
systems. For this reason, it has the potential to play an important role in
the imaging of angiogenesis. The advantages of sonography in this context are
the same as those in general clinical practice—that is, low financial
cost, portability, lack of necessity for contrast material, and lack of
restrictions in performing frequent serial examinations at short intervals.
Poor specificity (i.e., inability to accurately distinguish tumoral blood
vessels from normal vessels) is one limitation of sonography for the depiction
of angiogenesis. However, this disadvantage can be overcome by using
appropriate angiogenesis-targeted contrast agents, which are outlined in the
section titled Direct Imaging of Angiogenesis.
Doppler sonography has been used to great advantage to depict tumor
vascularity in the clinical setting and, in some cases, to provide prognostic
information about the likelihood of metastasis based on vascularity
[41]. However, studies
correlating sonographic features of tumors with markers of angiogenesis in
humans are relatively few. In one study, investigators correlated peak
systolic velocity in ovarian tumors with microvessel density and expression of
an angiogenic peptide in resected specimens
[42]. Investigators have also
analyzed spectral waveforms in tumors to determine whether the resistive index
correlates with angiogenic features. Both low resistive indexes (thought to
reflect diminished vasomotor control and arteriovenous shunting) and high
resistive indexes (possibly reflecting high interstitial pressure due to
microvascular permeability) have been reported in tumors, thereby limiting the
specificity of these findings
[43]. In one study of cervical
carcinomas, investigators found a correlation between a lower resistive index
in vivo and higher microvascular density and a form of VEGF in biopsy
specimens [44].
Microbubble-enhanced color Doppler techniques are also reported to provide
substantially increased sensitivity and specificity for the imaging of
angiogenesis in neoplasms compared with other sonographic techniques. The use
of targeted microbubbles is discussed later. The reader is referred to an
excellent review article outlining the status of microbubble techniques and
other sonographic techniques for angiogenesis imaging
[43].
Novel Imaging Techniques
As established methods are being explored for use in angiogenesis imaging,
new techniques are also being brought to bear. Although these techniques hold
substantial promise, presently they are primarily used for animal model
studies and may not be available in the clinical setting for many years, if at
all. For instance, investigators have explored the use of photoacoustic
imaging of angiogenesis [45].
Photoacoustic imaging operates on the principle that light can be converted to
sound by optical absorption. Pulsed laser light can be percutaneously applied
and absorbed by hemoglobin molecules. The absorption, in turn, causes a small
degree of heating, some degree of dilation of blood vessels, and local
pressure elevations that generate ultrasonic waves
[45] (Figs.
3A and
3B). When measured at enough
locations, these ultrasonic waves can be used to provide a 3D reconstruction
of vessel position and to record increases in vascular density. Investigators
have effectively used this technique to quantitatively monitor angiogenesis in
superficial tumors grown in animal models
[45] (Figs.
3A and
3B).
Direct Imaging of Angiogenesis
In addition to indirect imaging of angiogenesis, one may also directly
depict angiogenesis using imaging agents targeted at proteins or receptors
involved in angiogenesis. One line of investigation centers on membrane
proteins that angiogenic vessels selectively express; a multiplicity of such
targets have been identified, including
ß3 integrin, prostate-specific membrane
antigen, thrombospondin-1 receptor, and VEGF and its membrane receptors
[46]. Various imaging
techniques are being explored for this purpose, with each technique having
relative advantages and disadvantages. For instance, MRI provides much better
spatial resolution than PET, but the concentration at which most molecular
targets are present in living tissues is in the picomolar to low micromolar
range, which is far below that which can be detected using conventional MRI
[47]. Despite its relatively
poor spatial resolution compared with MRI, PET is more sensitive to small
concentrations of radiolabeled agents.
View larger version (168K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7A —Direct imaging of angiogenesis using IV infusion of
endothelial precursor cells labeled with superparamagnetic particles 2 days
after intracerebral implantation of glioma cells in mice. (Reprinted with
permission from Anderson SA, Glod J, Arbab AS, et al. Noninvasive MRI of
magnetically labeled stem cells to directly identify neovasculature in a
glioma model. Blood 2005; 105:420-425
[54].) Coronal 3D RARE image
obtained 11 days after tumor implantation shows migration of radiolabeled
cells into tumor periphery (arrow).
|
|
View larger version (137K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7B —Direct imaging of angiogenesis using IV infusion of
endothelial precursor cells labeled with superparamagnetic particles 2 days
after intracerebral implantation of glioma cells in mice. (Reprinted with
permission from Anderson SA, Glod J, Arbab AS, et al. Noninvasive MRI of
magnetically labeled stem cells to directly identify neovasculature in a
glioma model. Blood 2005; 105:420-425
[54].) Ex vivo coronal
gradient-echo image shows rim of hypointense signal caused by deposition of
radiolabeled endothelial precursor cells in tumor margins.
|
|
Sonography
Various sonographic techniques have been advanced for the direct depiction
of angiogenesis, primarily through the use of novel contrast agents such as
gas-filled microbubbles, which have high backscatter signals
[48]. Such microbubbles can be
directed against specific targets such as endothelial cell receptors. Because
the microbubbles remain in the vasculature and nontargeted microbubbles
(depending on dosage) can be removed via the reticuloendothelial system within
an hour, a high signal-to-background ratio can be achieved
[47]. Another advantage is
that such contrast agents can be used on repeated occasions in relatively
short succession [47].
Microbubbles can be specifically used for angiogenesis imaging by directing
them against activated endothelium. This feature has been accomplished by
targeting the microbubbles against ß3
integrin, which is an epitope found on the luminal surface of angioblasts.
Conjugation of the microbubbles with a specific protein (i.e., echistatin)
derived from viper venom confers them high specificity for binding against
ß3 integrin
[49]. The resultant contrast
agent has been successfully used to image angiogenesis in both the setting of
atherosclerotic disease and in tumors in mouse models
[49,
50]. For instance,
investigators have used such microbubbles to depict neovasculature in
malignant gliomas grown in rats
[49] (Figs.
4A,
4B, and
4C). When
ß3 integrin-targeted microbubbles were
infused 28 days after tumor implantation, the microbubbles were seen to
selectively deposit in the tumors. Confocal microscopy further showed that the
microbubbles were preferentially sequestered in the intravascular space and
particularly in small neovessels (Figs.
4A,
4B, and
4C). Furthermore, postmortem
immunohistochemical staining showed good correlation between regions of
microbubble deposition and -integrin expression on the microvascular
endothelium of tumor neovessels.
View larger version (84K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8 —PET direct imaging of angiogenesis uses radiotracer composed
of monoclonal antibody that binds to human vascular endothelial growth factor
(VEGF) labeled with a positron-emitting radionuclide, iodine-124. Imaging of
tumor-bearing mouse in coronal (left), sagittal (middle),
and transverse (right) planes shows focal accumulation of antibody in
tumor. (Reprinted with permission from Collingridge DR, Carroll VA, Glaser M,
et al. The development of [(124)I]iodinated-VG76e: a novel tracer for imaging
vascular endothelial growth factor in vivo using positron emission tomography.
Cancer Res 2002; 62:5912-5919
[56].)
|
|
View larger version (94K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 9A —Use of rotational microCT to perform direct imaging of
angiogenesis in ex vivo specimens in rabbit tumor model. (Reprinted with kind
permission of Springer Science and Business Media from Maehara N. Experimental
microcomputed tomography study of the 3D microangioarchitecture of tumors.
Eur Radiol 2003; 13:1559-1565
[57]. MicroCT scan 1 day after
tumor implantation shows small cluster of abnormal vessels (arrow) at
tumor site.
|
|
View larger version (73K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 9B —Use of rotational microCT to perform direct imaging of
angiogenesis in ex vivo specimens in rabbit tumor model. (Reprinted with kind
permission of Springer Science and Business Media from Maehara N. Experimental
microcomputed tomography study of the 3D microangioarchitecture of tumors.
Eur Radiol 2003; 13:1559-1565
[57]. On day 3 after tumor
implantation in another rabbit than that shown in A, a larger
conglomeration of abnormal arteries and veins (arrow) than shown in
A is seen at tumor site.
|
|
View larger version (119K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 9C —Use of rotational microCT to perform direct imaging of
angiogenesis in ex vivo specimens in rabbit tumor model. (Reprinted with kind
permission of Springer Science and Business Media from Maehara N. Experimental
microcomputed tomography study of the 3D microangioarchitecture of tumors.
Eur Radiol 2003; 13:1559-1565
[57]. Seven days after tumor
implantation in a third rabbit, clusters of tortuous, dilated vessels are seen
at margins of tumor (arrowheads), and new small vessels
(arrow) are seen centrally.
|
|
MRI
A number of MRI techniques have been developed for the direct imaging of
angiogenesis. In one study, investigators studied angiogenesis in neointimal
proliferation in an animal model of atherosclerotic plaque development
(expected to stimulate angiogenesis)
[51]. The investigators
infused paramagnetic nanoparticles targeted against the aforementioned
ß3 integrin (which is not present on
mature endothelial cells). These nanoparticles were modified for targeting of
ß3 by coating the surface of each
nanoparticle with approximately 90,000 paramagnetic gadopentetate
dimeglumine-based chelates. The investigators then performed T1-weighted
imaging that depicted selective binding of the paramagnetic nanoparticles at
sites of the adventitia of the aortic wall that showed marked proliferation of
angiogenic vessels (Figs. 5A,
5B, and
5C).
Using a rabbit tumor model, investigators have also used
ß3 integrin-targeted nanoparticles to
depict tumoral angiogenesis using T1-weighted MRI
[52]
(Fig. 6). Reportedly, the
paramagnetic nanoparticles penetrate into immature (leaky) vessels and
extravasate locally but do not diffuse into the interstitial space. Deposition
of paramagnetic particles was seen predominantly in the tumor periphery.
However, in other studies using different size nanoparticles and other types
of tumor implantation, diffuse infiltration throughout the entire tumor has
been reported [53].
Yet another method for direct imaging of angiogenesis using MRI involves
the IV infusion and tracking of subsequent migration of labeled endothelial
precursor cells into tumors (Figs.
7A and
7B). Understanding the rate of
incorporation of these cells into tumor vasculature could provide a
measurement of tumor angiogenesis. In one study, investigators labeled bone
marrow-derived endothelial precursor cells with superparamagnetic particles
and followed their migration into the tumor periphery using MRI
[54]. They found that,
approximately 9 days after infusion, hypointense regions on gradient-echo
images consistent with cell migration into the tumor vasculature were
seen.
PET
PET holds substantial promise for use in the direct imaging of angiogenesis
in animal (and perhaps, eventually human) models. As one of many examples, PET
radiotracers can be bound to antibodies directed against factors associated
with angiogenesis, such as angiogenesis-promoting growth factors or receptors
for such growth factors. In particular, attention has focused on the thymidine
kinases, which are transmembrane receptors that mediate VEGF activity.
Antibodies against such growth factors are presently being used in
experimental trials for the treatment of tumors; these antibodies can be
labeled with radionuclides such as technetium-99m, iodine-124, and
samarium-153 [55,
56]
(Fig. 8).
MicroCT
MicroCT is a recently developed imaging tool that can provide
very-high-resolution images (
50- to 100-µm resolution) of body parts
in small animals such as mice. High-resolution vascular imaging can be
accomplished after the IV administration of CT contrast material. In small
animal models, this technique has allowed serial imaging of vessel number and
morphology in ex vivo specimens, which might be useful in the evaluation of
angiogenesis and the response to therapy in experimental models. In one study
that used rotational microCT to assess neovascularization of carcinomas in
rabbits, new blood vessels having a diameter of approximately 50 µm were
detected within the first few days after tumor transplantation (Figs.
9A,
9B, and
9C); serial imaging across the
group of animals allowed determination of serial changes in vessel number and
morphology [57]. Although not
applicable in humans, techniques of this type could potentially be of value in
the preclinical assessment of antiangiogenic agents.
Optical Imaging
The development of optical imaging techniques has provided numerous new
inroads to the depiction of tumor physiologic processes in general and
angiogenesis in particular. Bioluminescence imaging—that is, detection
of bioluminescent light from internal sources in living tissues (which can be
accomplished in some scenarios independently of external excitation for
emission of luminescence)—is one of many optical imaging techniques that
has been used for the assessment of angiogenesis
[58,
59].
In one study, investigators used a window chamber model (in which a tumor
is grown in the flank of a mouse and directly visualized by means of a
transparent window placed over the growing tumor) and expression of reporter
genes to study angiogenesis
[60]. Those investigators
found that the presence of tumor cells alters the morphology of preexisting
blood vessels just a few days after tumor implantation. Using tumors cells
that expressed green fluorescent protein, whose location and number could
therefore be visually tracked using the window chamber, these investigators
saw that tumor cells migrated toward preexisting blood vessels, suggesting a
chemotaxislike migration of tumor cells in response to signaling mechanisms
from blood vessels before angiogenesis. The tumor cells were then seen to grow
along and around blood vessels.
In another study using fluorescence imaging rather than bioluminescence
imaging, investigators using molecular probes that are photoactive in the
near-infrared range (Fig. 10)
were able to depict distribution of ß3
integrin (implicated in angiogenesis, as discussed previously)
[61]. Studies performed ex
vivo showed that probe accumulation in tumors correlated with mitochondrial
NADH (nicotinamide adenine dinucleotide) concentration, suggesting that
angiogenesis (as marked by expression of
ß3 integrin) also correlates with
metabolic status.
Summary
Therapies directed against tumor angiogenesis are developing at a rapid
pace. At present, the best method for assessing effects of these therapies in
humans is a matter of active discussion among investigators. Because most of
the surrogate markers for assessment are imaging-based, it is important that
radiologists understand the type of questions being asked by investigators and
clinicians using these therapies. Many intriguing methods for directly imaging
angiogenesis have been developed for preclinical studies in animals. It is
hoped that these imaging techniques will have applications in humans. However,
even if that does not prove to be so, these novel imaging techniques are
likely to help in understanding the effects of antiangiogenic agents and to
decrease the time to implementation of these therapies in humans.
References
- Lip GY, Blann AD. Thrombogenesis, atherogenesis and angiogenesis in
vascular disease: a new "vascular triad." Ann
Med 2004;36:119
-125[CrossRef][Medline]
- Folkman J. Tumor angiogenesis: therapeutic implications.
N Engl J Med1971; 285:1182
-1186[Medline]
- Gaya AM, Rustin GJ. Vascular disrupting agents: a new class of drug
in cancer therapy. Clin Oncol2005; 17:277
-290[CrossRef]
- National Cancer Institute. Angiogenesis inhibitors in
clinical trials. Available at:
www.cancer.gov/clinicaltrials/developments/anti-angio-table.
Accessed February 23, 2006
- Kan Z, Phongkitarun S, Kobayashi S, et al. Functional CT for
quantifying tumor perfusion in antiangiogenic therapy in a rat model.
Radiology2005; 237:151
-158[Abstract/Free Full Text]
- Ellenberg SS, Hamilton JM. Surrogate endpoints in clinical trials:
cancer. Stat Med1989; 8:405
-413[Medline]
- Galbraith SM. Antivascular cancer treatments: imaging biomarkers in
pharmaceutical drug development. Br J Radiol2003; 76:S83
-S86[Free Full Text]
- Persigehl T, Heindel W, Bremer C. MR and optical approaches to
molecular imaging. Abdom Imaging2005; 30:342
-354[Medline]
- Padhani AR, Leach MO. Antivascular cancer treatments: functional
assessments by dynamic contrast-enhanced MRI. Abdom
Imaging 2005;30:324
-341[Medline]
- Miller JC, Pien HH, Sahani D, Sorensen AG, Thrall JH. Imaging
angiogenesis: applications and potential for drug development. J
Natl Cancer Inst 2005;97:172
-187[Abstract/Free Full Text]
- Cha S. Perfusion MRI of brain tumors. Top Magn Reson
Imaging 2004;15:279
-289[CrossRef][Medline]
- Provenzale JM, Mukundan S, Barboriak DP. Diffusion-weighted and
perfusion MR imaging for brain tumor characterization and assessment of
treatment response. Radiology2006; 239:632
-649[Abstract/Free Full Text]
- Provenzale JM, Wang GR, Brenner T, Petrella JR, Sorensen AG.
Comparison of permeability in high-grade and low-grade brain tumors using
dynamic susceptibility contrast MRI. AJR2002; 178:711
-716[Abstract/Free Full Text]
- Pham CD, Roberts TP, van Bruggen N, et al. MRI detects suppression
of tumor vascular permeability after administration of antibody to vascular
endothelial growth factor. Cancer Invest1998; 16:225
-230[Medline]
- Preda A, Novikov V, Moglich M, et al. MRI monitoring of Avastin
antiangiogenesis therapy using B22956/1, a new blood pool contrast agent, in
an experimental model of human cancer. J Magn Reson
Imaging 2004;20:865
-873[CrossRef][Medline]
- Rugo HS, Herbst RS, Liu G, et al. Phase I trial of the oral
antiangiogenesis agent AG-013736 in patients with advanced solid tumors:
pharmacokinetic and clinical results. J Clin Oncol2005; 23:5474
-5483[Abstract/Free Full Text]
- Figg WD, Dahut W, Duray P, et al. A randomized phase 2 trial of
thalidomide, an angiogenesis inhibitor, in patients with androgen-independent
prostate cancer. Clin Cancer Res2001; 7:1888
-1893
Top
Abstract
Introduction
