PET Imaging of Angiogenesis

Synopsis

Angiogenesis is a highly-controlled process that is dependent on the intricate balance of both promoting and inhibiting factors, involved in various physiological and pathological processes. A comprehensive understanding of the molecular mechanisms that regulate angiogenesis has resulted in the design of new and more effective therapeutic strategies. Due to insufficient sensitivity to detect therapeutic effects by using standard clinical endpoints or by looking for physiological improvement, a multitude of imaging techniques have been developed to assess tissue vasculature on the structural, functional and molecular level. Imaging is expected to provide a novel approach to noninvasively monitor angiogenesis, to optimize the dose of new antiangiogenic agents and to assess the efficacy of therapies directed at modulation of the angiogenic process. All these methods have been successfully used preclinically and will hopefully aid in antiangiogenic drug development in animal studies. In this review article, the application of PET in angiogenesis imaging at both functional and molecular level will be discussed. For PET imaging of angiogenesis related molecular markers, we emphasize integrin α_vβ₃, VEGF/VEGFR, and MMPs.

Introduction

Angiogenesis refers to the process by which new blood vessels are formed and is involved in various physiological as well as pathological processes including physical development, wound repair, reproduction, response to ischemia, arthritis, psoriasis, retinopathies, solid tumor growth and metastatic tumor spread¹. Angiogenesis is a highly-controlled process that is dependent on the intricate balance of both promoting and inhibiting factors.

Antiangiogenic and antivascular agents are intensively investigated for tumor therapy and can be potentially used to control eye diseases and arthritis. Pro-angiogenic therapy are also undergoing clinical trials in human patients suffering from ischemic heart disease, peripheral vascular disease, chronic wounds or stroke². A comprehensive understanding of the molecular mechanisms that regulate angiogenesis has resulted in the design of new and more effective therapeutic strategies³. Although preclinical animal studies have demonstrated the benefit of pro-angiogenic therapy, recent clinical trials focused on the stimulation of myocardial or peripheral angiogenesis by the local delivery of growth factors were somewhat disappointing, showing no clear benefit over placebo in patients with severe ischemia⁴. Most of these studies evaluated angiogenesis by using standard clinical endpoints or by looking for a physiological improvement (i.e. improvement in perfusion). These approaches may have insufficient sensitivity to detect a therapeutic benefit.

For tumor therapy, Bevacizumab, a humanized monoclonal antibody directed against vascular endothelial growth factor (VEGF), is the first drug developed as an inhibitor of angiogenesis approved by the Food and Drug Administration (FDA)^5–7. Sorafenib and sunitinib that target multiple receptor tyrosine kinases (VEGF receptors and platelet-derived growth factor (PDGF) receptors), have also been approved by the FDA as antiangiogenic drugs⁸. Traditionally, the gold standard to evaluate therapeutic response is tumor volume change. Clinical trials with conventional cytotoxic chemotherapeutic agents have mainly used morphological imaging to provide indices of therapeutic response, mostly computed tomography (CT) or magnetic resonance imaging (MRI) according to the Response Evaluation Criteria in Solid Tumors (RECIST) introduced in the year 2000⁹. However, antiangiogenic agents are typically cytostatic rather than cytotoxic and lead to a stop or delay of tumor progression rather than tumor shrinkage. Thus, it is not sensitive to use tumor volume as an indicator for therapeutic efficacy evaluation and it might take months or years to assess.

Both the success and setback in angiogenesis related therapies spur the need for the development of noninvasive imaging strategies for the direct noninvasive evaluation of molecular events associated with angiogenesis. Imaging is expected to provide a novel approach to noninvasively monitor angiogenesis, to optimize the dose of new antiangiogenic agents and to assess the efficacy of therapies directed at modulation of the angiogenic process^10,11–13. In this review article, after brief introduction of angiogenesis biology and structure and functional imaging of angiogenesis with various imaging modalities, we will focus on the application of PET in angiogenesis imaging at both functional and molecular level.

Biology of Angiogenesis

The whole angiogenesis process involves several steps including the growth of endothelial sprouts from preexisting postcapillary venules and following the growth and remodeling process of the primitive network into a complex network¹⁴. The cellular and molecular mechanisms of angiogenesis differ in various tissues and physiologic or pathologic angiogenesis¹⁵. Here we will give a brief introduction of tumor angiogenesis.

Each solid malignancy starts as a small population of transformed cells, which do not initially have a blood supply of their own. Tumor cells are initially supplied by diffusion and tumor growth is limited by the lack of access to growth factors, circulating oxygen, and nutrients¹⁶. Without angiogenesis, the growth of solid tumors remains restricted to 2–3 mm in diameter¹⁷. Tumor angiogenesis occurs as a series of events^{18, 19}. First, diseased tissues produce and release angiogenic growth factors that diffuse into the nearby tissues in response to tumor hypoxia, such as the vascular endothelial growth factor (VEGF), the acidic and basic fibroblast growth factors (aFGF, bFGF), and the platelet-derived endothelial cell growth factor (PD-ECGF)²⁰. When the angiogenic growth factors bind to their corresponding specific receptors located on the endothelial cells of pre-existing blood vessels, various signal transduction pathways are activated, for example phosphorylation of tyrosine kinases, protein kinases, and MAP kinases and consequently to the activation of endothelial cells^{21, 22}. Consequently, the original vessels undergo characteristic morphological changes, including enlargement of the diameter, basement membrane degradation, a thinned endothelial cell lining, increased endothelial number, decreased pericyte number and pericyte detachment²³. In the next step, several different mechanisms may lead to the formation of new tumor blood vessels^{24, 25}. The original vessels may retain their large diameter and evolve into medium-sized arteries and veins by acquiring a smooth muscle and internal elastica. Alternatively, the endothelium of a mother vessel may form smaller separate well-differentiated vessel channels by projecting cytoplasmic structures into the lumen which form translumenal bridges. A third process is called intussusception and involves focal invagination of connective tissue pillars from within the mother vessel. Finally, endothelial cell sprouting may occur, which requires the focal dissolution of the basement membrane surrounding mother vessels²⁶. This is achieved by a number of proteolytic enzymes, including matrix metalloproteinases (MMPs) and plasminogen activator, which enable endothelial cells to exit the vessel. Activated angiogenic endothelial cells proliferate rapidly and migrate into the extracellular matrix towards the angiogenic stimulus²⁷. Cell surface adhesion molecules such as integrins play an important role in endothelial cell migration and in contact with the extracellular tumor matrix, facilitate cell survival^{28, 29}. At the sprouting tips of growing vessels, endothelial cells secrete MMPs that facilitate degradation of the extracellular matrix and cell invasion³⁰. Next, a lumen within an endothelial cell tubule has to be formed, which requires interactions between the extracellular matrix and cell-associated surface proteins, among them are galectin-2, PECAM-1, and VE-cadherin³¹. Finally, newly formed vessels are stabilized through the recruitment of smooth muscle cells and pericytes.

Structural Imaging of Vasculature/Angiogenesis

All imaging modalities can provide structural information although they have different spatial resolution. The old-fashion way for vascular structure imaging is x-ray angiography. However, it is difficult to provide microvasculature information. Following the steps of improvement of imaging equipments, contrast agents and data acquisition and analysis techniques, more detailed vascular structure was deciphered. Several modalities are available for tumor microvascular imaging, including intravital microscope, CT angiography, contrast enhanced ultrasound and high-resolution magnetic resonance angiography (MRA)³². Ex vivo structural imaging of tumor vasculature can be achieved by various techniques such as vascular casts^{33, 34}, immunohistochemical staining of endothelial cell markers such as CD31 and von Willebrand factor^{35, 36}, labeling the endothelial cells by fluorescent reporters expressed in transgenic mice^{37, 38}, and intravital labeling^{39, 40}. Tumor macrovasculature imaging can be performed clinically by various imaging modalities such as computed tomography (CT)^{41, 42}, magnetic resonance imaging (MRI)^{43, 44}, and ultrasound⁴⁵. However, visualization of the microvasculature is very challenging even after administration of intravascular contrast agents. Scanners dedicated to small animal imaging studies, such as microCT, have better spatial resolution in pre-clinical models but with poor temporal resolution and large radiation exposure ⁴⁶.

Intravital microscopy of tumors growing in window chambers in animal models can directly investigate tumor angiogenesis and vascular response to treatment, in terms of both the morphology of the vascular networks and the function of individual vessels ⁴⁷. This technique allows for repeated measurements of the same tumor with very high resolution (down to sub-micrometer level). Multi-photon fluorescence microscopy techniques have also been applied to these model systems to obtain 3D images of the tumor vasculature⁴⁷.

Functional Imaging of Vasculature

The major consequence of angiogenesis is to perfuse and oxygenate surrounding tissue; therefore, the angiogenic process can be assessed by the evaluation of standard physiologic parameters such as regional perfusion, function, and metabolism. During antiangiogenic or pro-angiogenic therapies, the changes in hemodynamic parameters can also be promising biomarkers for evaluating the therapeutic effect along with morphological changes. Traditionally, tumor angiogenesis and anti-angiogenic therapy have been evaluated by methods such as measurement of circulating angiogenic markers and histological estimate of microvascular density (MVD). Various imaging modalities including dynamic contrast enhanced MRI (DCE-MRI), ultrasound, PET (especially with [¹⁵O]water), dynamic contrast enhanced CT are currently employed to provide functional information of the vasculature⁴⁸.

DCE-MRI has been well established to investigate angiogenesis within tumors, and in particular the response to antiangiogenic therapy. DCE-MRI works by tracking the pharmacokinetics of injected contrast agents as they pass through the tumor vasculature, which represents a complex summation of vascular permeability, blood flow, vascular surface area, and interstitial pressure^{49, 50}.

DCE-MRI can be performed with low-molecular-weight contrast media (LMCM) such as Gd-diethylenetriamine pentaacetic acid (Gd-DTPA) or macromolecular contrast media (MMCM) such as Gd conjugated human serum albumin (Gd-HSA)⁵¹. It has been shown that DCE-MRI can detect responses to PTK/ZK (a VEGF receptor tyrosine kinase inhibitor) therapy as early as two days after therapy with significant reductions in area under gadolinium-contrast-medium curve (AUGC)⁵² or permeability parameters⁵³, which also predict subsequent response. LMCM DCE-MRI has also shown significant reductions in permeability values in patients treated with the antivascular agents AG-013736 (an inhibitor of the VEGF, PDGF, and c-Kit receptor tyrosine kinases) and SU5416 (a selective inhibitor of VEGFR-2 tyrosine kinase) activity⁵⁴. Although consensus is still lacking on the exact kinetic model to be used in analyzing DCE-MRI data, the differences among the various methods are often marginal. Therefore, DCE-MRI is rapidly emerging as the imaging technique of choice for monitoring clinical response in trials of new antiangiogenic and antivascular therapies. Unlike LMCM, the increased size of MMCMs makes them less diffusible, and K^trans values may reflect permeability within tumors more accurately⁵⁵. MMCMs can also give more accurate estimates of tumor blood volume since they are excellent blood pool agents. For example, SU6668 is an oral, small molecule inhibitor of angiogenic receptor tyrosine kinases such as vascular endothelial growth factor receptor 2 (Flk-1/KDR), platelet-derived growth factor (PDGF) receptor, and fibroblast growth factor (FGF) receptor. DCE-MRI clearly detected the early effect (after 24 h of treatment) of SU6668 on tumor vasculature as a 51% and 26% decrease in the average vessel permeability measured in the tumor rim and core, respectively. A substantial decrease was also observed in average fractional plasma volume in the rim (59%) and core (35%) of the tumor^{56, 57}. In addition to DCE-MRI, other MRI techniques have also been developed to retrieve functional information of the vasculature. In arterial spin labeling (ASL), water molecules can be labeled for MRI by inverting the nuclear spin of their hydrogen atoms with a radiofrequency pulse directed at the arterial blood before it enters the regions of interest (ROIs)^{58, 59}. An absolute value of blood flow is determined by the change in the MR signal as the labeled water in the arterial bloodstream arrives in the ROI⁶⁰. Blood oxygen level dependent (BOLD) MRI can detect the changes in oxygen saturation of the blood and this effect can be enhanced by increasing the amount of oxygen in the breathed air^{61, 62}.

Ultrasound is also well established as a means of measuring blood flow or, more precisely, blood velocity using the Doppler principle or microbubles as contrast agent^63–66. Power Doppler can be quantified to give an estimate of the relative fractional vascular volume while microbubbles can show blood flow down to the microcirculation level by raising the signal from smaller vessels^{67, 68}. Specialized contrast-specific US techniques have been developed for improving image qualities such as Pulse Inversion^{69, 70} and Power Modulation⁷¹. Given the fundamental assumption that the relation between microbubble concentration video-intensity is linear up to the achievement of a plateau phase⁷², in animal models CEUS can quantify tumor vascularity determined by neoangiogenesis^{73, 74}. The use of ultrasound contrast agents and nonlinear processing provide access to the bulk properties of the microvascular compartment but they do not offer sufficient resolution to observe the morphology and detailed flow characteristics of the microvasculature. With high frequency ultrasound, it is possible to achieve resolution ranging from 15 to 100 μm in the 20–100MHz range⁴⁵. However, it is subject to an inherent trade-off between image resolution and imaging depth⁷⁵. Ultrasound (particularly microbubble contrast enhanced ultrasound) is a valuable imaging modality to determine the tumor microvascular blood volume and blood velocity⁷⁶. Dynamic contrast-enhanced ultrasonography (DCE-US) allows repeated examinations and provides both morphologic and functional analyses. US modes, based on the second harmonic signal generated by the nonlinear properties of contrast agents, have provided access to tumor blood flow with the quantification of the contrast-uptake kinetics within tumors after a bolus injection of contrast agent⁷⁷. Several quantitative parameters considered as indicators of tumor flow such as the peak intensity (PI) or time-to-PI can be extracted from the time-intensity curves of contrast uptake⁷⁸. Using DCE-US, the antitumor efficacy of AVE8062, a tumor vasculature disruptive agent, has been assessed in melanoma-bearing nude mice⁷⁹.

Many other imaging modalities can also reveal the functional properties of the vasculature. DCE-CT is analogous to DCE-MRI^{80, 81}. To minimize the exposure to ionizing radiation and the nephrotoxicity of CT contrast agents, DCE-CT studies are typically quite brief using only a low dose of contrast agent. Optical imaging can also be applied to evaluate important functional indexes of blood vessels such as vascular permeability, vessel size, and blood flow⁸². Multiphoton microscopy in combination with fluorescently labeled molecules can be used to quantify the permeability of individual tumor blood vessels non-invasively deep inside living animals⁸³. Fluorescence-mediated tomography (FMT) has been applied to measure angiogenesis in superficial tissue by using fluorescent nanoparticles⁸⁴. However, optical imaging is still limited to tissue and animal models.

PET Imaging

So far, PET is the most sensitive and specific technique for imaging molecular pathways in vivo in humans⁸⁵. PET radiotracers are physiologically and pharmacologically relevant compounds labeled with positron-emitting radioisotopes (such as fluoride-18 or carbon-11). After internalization by injection or inhalation, the tracer reaches the target and the location and the quantity is then detected with a PET scanner. With a ring-shaped array of photoelectric crystals, PET detectors capture “coincidentally” a pair of 511 keV photons at almost 180° separation emitted by interaction of a positron with negatively charged electrons. The raw PET scan data are the set of coincidental photoelectric events, logged for time and location, which indicate the position of the molecule spatiotemporally. Using reconstruction algorithms, images can then be constructed tomographically, and regional time activities can be derived⁸⁶.

The inherent sensitivity and specificity of PET is the major strength of this technique. Isotopes can be detected down to the 100 picomolar level in the target tissues. At this low level the compounds often have little or no physiological effect on the patient or the test animal, which permits studying the mechanism of action or biodistribution independent of any physiological consequences^{87, 88}. The spatial resolution of PET down to the millimeter level permits applications not only to humans for diagnosis and drug development, but also to animals for preclinical studies. The ability of PET to translate studies from animals to humans adds to its appeal.

Compared with single photon emission computed tomography (SPECT), PET offers increased spatial information and permits more accurate attenuation correction. Many PET radiotracers have a short half-life, which allows for repetitive imaging over time. However, the anatomical resolution of PET (approximately 4–8 mm³ in clinical and 1–2 mm³ in small animal imaging systems) is noticeably poorer than that achieved by CT or MRI⁸⁹. The variable movement of positrons before annihilation, and the deviation of the generated 511 keV photons from the exact 180° angular separation, can limit the resolution. To overcome this limitation, hybrid systems such as PET-CT have been introduced⁹⁰. The CT component of the hybrid system is used to improve anatomical definition of the ROIs for analysis and to create radiation-attenuation maps to correct for non-uniform attenuation⁹¹. With the development of microPET or microPET/CT scanners dedicated to small animal imaging studies, it can provide a similar in vivo imaging capability in mice, rats, monkeys, and humans so one can readily transfer knowledge and molecular measurements between species^{92, 93}. Initial experiments with PET/MRI prototypes also showed very promising results, indicating its great potential for clinical and preclinical imaging⁹⁴.

Functional Imaging of Angiogenesis with PET

A major advantage of the nuclear medicine techniques especially using PET tracers is that they are truly quantitative and that the tissue concentration C_t can be measured non-invasively⁹⁵. ¹³³Xe ⁹⁶ has been used to measure regional cerebral blood flow and ¹¹C-microspheres of approximately 10 μm diameter been used as “gold standard” for perfusion measurements or for validation of new imaging methods for perfusion measurement ⁹⁷.

Nowadays, most PET-perfusion measurements are performed using ¹⁵O-H₂O, using either static or dynamic PET imaging^{98, 99}. ¹⁵O-H₂O satisfies all the requirements for a perfusion tracer in Fick’s model¹⁰⁰ (1) because it is biologically and metabolically inert, and freely diffusible into and out of tissue water.

$$C_t=P\ast \int (C_i-C_e)d_t$$ (1)

C_t = tissue concentration, mol*ml_tissue⁻¹; C_i = Influx concentration, mol*ml_carrier⁻¹

C_e = eflux concentration, mol*ml_carrier⁻¹; P = Perfusion, ml_carrier * min⁻¹* ml_tissue⁻¹

Thus “tissue water” can be modelled as a single compartment including both tissue and its draining fluids (lymphatics and veins). Two methods can be used for measuring perfusion with [¹⁵O]H₂O, the steady-state method of Frackowiak and colleagues, and the ¹⁵O-dynamic water method by Lammertsma et al.¹⁰¹. The latter is currently used most often for perfusion studies due to improved PET scanner technology. The tracer is administered by inhalation or by peripheral venous bolus injection. Continuous arterial data are obtained either by image based arterial input functions (a large vessel like the aorta or the left ventricle) or by peripheral sampling to a well counter device. The data are compatible with those from diseases investigated with other methods, and the values reported by PET for tumors are within the reported range for PET in other tissues^{102, 103}. In locally advanced breast cancer, first results with dynamic ¹⁵O-H₂O PET are promising, as tumor blood flow decreased in the responder group after chemotherapy, whereas it increased in the non-responder group¹⁰⁴.

PET imaging can also be used to derive data on blood volume and vascular permeability. Blood volume imaging with PET uses ¹⁵O-CO or ¹¹C-CO carbon monoxide. ¹⁵O-CO binds irreversibly with hemoglobin to form ¹⁵O-CO-Hb carboxyhemoglobin ⁸⁶. Because ¹⁵O-CO-Hb remains exclusively within the vasculature it can be used as a tracer of vascular volume. A tissue concentration dataset is obtained over a further 5–6 min and an arterial ¹⁵O-CO-Hb concentration curve is derived from a series of arterial blood samples over the same interval. Another method for blood volume imaging is labeling red blood cells (RBCs) or albumin with radionuclides, because both are too large to leave normal blood vessels and are retained in the blood pool. In tumor vessels, leakage of these contrast agents into the tumor will occur, but this effect can be used to calculate the tumor vessel permeability, when dynamic imaging is performed. For PET, the tracer ⁶⁸Ga-DOTA-albumin has been developed and showed favorable results in first animal studies¹⁰⁵.

Imaging of molecular markers of vasculature

Even though structural/functional imaging of the vasculature can reveal potentially useful information before, during, and after therapeutic intervention, they do not convey enough knowledge about the biological changes upon therapy at the molecular level which may occur long before any structural/functional changes can be detected. While techniques like DCE-MRI and ¹⁵O-H₂O PET for the assessment of hemodynamic parameters are widely used, the interpretation of the results with regard to their physiological meaning often remains difficult. Therefore more specific markers of angiogenic activity are necessary for pre-therapeutic assessment of angiogenesis and response evaluation during therapy. One approach is to identify molecular markers of angiogenesis such as receptors, enzymes, or extracellular matrix proteins and to use specific ligands to these targets conjugated with imaging probes for PET, SPECT, MRI, Optical Imaging or Ultrasound ^106-108. Several molecular imaging makers including integrins, VEGF/VEGFR, MMPs, Hypoxia/HIF1 are angiogenesis related and PET imaging targeting to these markers will be discussed in the following sections.

PET imaging of integrins

Integrins, a family of cell adhesion molecules, are involved in a wide range of cell-extracellular matrix (ECM) and cell-cell interactions^{28, 109}. Integrins are heterodimeric transmembrane glycoproteins consisting of different α- and β-subunits which play an important role in cell-cell- and cell-matrix-interactions¹¹⁰. In mammals, 18 α and 8 β subunits assemble into at least 24 different receptors¹¹¹. Integrins expressed on endothelial cells modulate cell migration and survival during angiogenesis while integrins expressed on carcinoma cells potentiate metastasis by facilitating invasion and movement across the blood vessels. The α_vβ₃ integrin, which binds to Arginine-Glycine-Aspartic acid (RGD)-containing components of the interstitial matrix such as vitronectin, fibronectin and thrombospondin^{112, 113}, is expressed in a number of tumor types such as melanoma, late stage glioblastoma, ovarian, breast, and prostate cancer^114–116. The critical role of integrin α_vβ₃ in tumor invasion and metastasis arises from its ability to recruit and activate MMP-2 and plasmin, which can degrade components of the basement membrane and interstitial matrix¹¹⁷. Among all 24 integrins discovered to date, integrin α_vβ₃ is the most intensively studied, though many other integrins such as α_vβ_1, α_vβ_5, α₅β₁ and α₄β₁ also play important roles in regulating angiogenesis^{29, 118–121}.

Several ECM proteins like vitronectin, fibrinogen and fibronectin interact via the RGD tripeptide sequence with the integrins¹¹³. Based on these findings, linear as well as cyclic RGD peptides have been introduced and showed high affinity and selectivity for α_vβ₃^{122, 123}. First in vivo application of radioiodinated RGD peptides revealed the receptor-specific tumor uptake but also predominantly hepatobiliary elimination, resulting in high activity concentration in the liver and small intestine¹²⁴. Consequently, several strategies to improve the pharmacokinetics of radiohalogenated peptides have been studied including conjugation with sugar moieties, hydrophilic amino acids and polyethylene glycol (PEG)^125–128. Besides radiohalogenated RGD peptides, a variety of radiometalated tracers have been developed as well, including peptides labeled with ¹¹¹In, ^99mTc, ⁶⁴Cu, ⁹⁰Y, ¹⁸⁸Re and ⁶⁸Ga ^129–132. Most of them are based on the cyclic pentapeptide and are conjugated via the γ-amino function of a lysine with different chelator systems, like DTPA, the tetrapeptide sequence H-Asp-Lys-Cys-Lys-OH, 1, 4, 7, 10-tetraazacyclododecane-N-N″-N″-N‴-tetraacetic acid (DOTA) and 1,4,7-triazacyclononane-1,4,7-triacetic acid (NOTA). While all these compounds have shown high receptor affinity and selectivity and specific tumor accumulation, the pharmacokinetics of most of them still have to be improved¹³³. Among them, the compound ^99mTc-NC100692 by GE Healthcare has been used for SPECT imaging in preclinical and clinical studies¹³⁴.

In a human melanoma M21 model, ¹⁸F-Galacto-RGD showed a tumor uptake of 1.5 %ID/g at 120 min p.i.^{135, 136}. Integrin receptor specific accumulation was demonstrated by blocking experiments injecting c(RGDfV) 10 min prior to tracer injection, which reduced tumor accumulation to approximately 35% of control. A correlation between integrin expression and tracer accumulation was observed in imaging studies with mice bearing melanoma tumors with increasing amounts of α_vβ₃-positive cells¹³⁷. These data demonstrate that non-invasive determination of α_vβ₃ expression and quantification with radiolabeled RGD peptides is feasible with PET scans. ¹⁸F-Galacto-RGD has also been applied to patients and successfully imaged α_vβ₃ expression in human tumors with good tumor/background ratios¹³⁸. Rapid clearance of ¹⁸F-Galacto-RGD from the blood pool and primarily renal excretion was confirmed by following biodistribution and dosimetry studies. Background activity in lung and muscle tissue was low and the calculated effective dose is very similar to an ¹⁸F-FDG scan¹³⁹ (Figure 1). Results from dynamic emission scans over 60 minutes and kinetic modeling studies suggested that SUVs derived from static emission scans at ~ 60 min. p.i. can be used for the assessment of α_vβ₃ receptor density with reasonable accuracy¹³⁹. SUVs and tumor/blood ratios based on PET imaging using ¹⁸F-Galacto-RGD were also found to correlate with the intensity of immunohistochemical staining of α_vβ₃ expression as well as with the microvessel density¹⁴⁰. Good tumor/background ratios with ¹⁸F-Galacto-RGD PET also have been demonstrated in squamous cell carcinoma of the head and neck (SCCHN) with a widely varying intensity of tracer uptake. Immunohistochemistry demonstrated predominantly vascular α_vβ₃ expression, thus in SCCHN, ¹⁸F-Galacto-RGD PET might be used as a surrogate parameter of angiogenesis¹⁴¹. Moreover, there was no obvious correlation between the tracer uptake of ¹⁸F-FDG and ¹⁸F-Galacto-RGD in patients with various tumors, indicating that α_vβ₃ expression and glucose metabolism are not closely correlated in tumor lesions and that consequently ¹⁸F-FDG cannot provide similar information as ¹⁸F-Galacto-RGD¹⁴².

Figure 1.

¹⁸F-Galacto-RGD scans of 2 patients with metastases from malignant melanoma and different tracer uptake. (Upper row images) An 89-y-old female patient with metastasis in subcutaneous fat in gluteal area on left side (arrow with dotted line). Tumor can be clearly delineated in CT scan (A), whereas it shows no significant uptake in ¹⁸F-Galacto-RGD PET scan (B; 60 min after injection). (Lower row images) A 36-y-old female patient with lymph node metastasis in right groin (arrow). Again, tumor is clearly visualized in CT scan (C) but also shows intense tracer uptake in ¹⁸F-Galacto-RGD PET scan (D; 89 min after injection; SUV, 6.8). Reproduced from Beer et al. ¹³⁹ with permission.

Within physiological ¹⁸F-Galacto-RGD uptake area, such as liver, spleen and intestine, lesion identification is still problematic. Therefore, multimeric RGD peptides have been developed in order to provide more effective antagonists with better targeting capability and higher cellular uptake through the integrin-dependent binding¹⁴³. The underlying rationale is that the interaction between integrin α_vβ₃ and RGD-containing ECM-proteins involves multivalent binding sites with clustering of integrins. A series of multimeric RGD peptides labeled with ¹⁸F or ⁶⁴Cu for PET imaging to improve the tumor-targeting efficacy and pharmacokinetics have been reported^{130, 144–148}. ¹⁸F-FB-E[c(RGDyK)]₂ (abbreviated as ¹⁸F-FRGD2) showed predominantly renal excretion and almost twice as much tumor uptake in the same animal model compared with the monomeric tracer ¹⁸F-FB-c(RGDyK)^{144, 145}. Tumor uptakes quantified by microPET scans in six tumor xenograft models correlated well with integrin α_vβ₃ expression level measured by SDS-PAGE autoradiography. The tetrameric RGD peptide-based tracer, ¹⁸F-E[E[c(RGDfK)]₂]₂, showed significantly higher receptor binding affinity than the corresponding monomeric and dimeric RGD analogues and demonstrated rapid blood clearance, high metabolic stability, predominant renal excretion and significant receptor-mediated tumor uptake with good contrast in xenograft-bearing mice ¹⁴⁸. Therefore, ¹⁸F-E[E[c(RGDfK)]₂]₂ is a promising agent for peptide receptor radionuclide imaging as well as targeted internal radiotherapy of integrin α_vβ₃ positive tumors. Compared with tetramer, RGD octamer further increased the integrin avidity by another 3-fold. In vivo microPET imaging showed that ⁶⁴Cu-DOTA-RGD octamer had slightly higher initial tumor uptake and much longer tumor retention in U87MG tumor that express high level of integrin¹⁴⁹ (Figure 2). However, compared with tetramers, higher renal uptake of the octamer was observed, which was attributed mainly to the integrin positivity of the kidneys. Wester and Kessler groups have also synthesized a series of monomeric, dimeric, tetrameric and octameric RGD peptides. These compounds contain different numbers of c(RGDfE) peptides connected via PEG linker and lysine moieties, which are used as branching units^{150, 151}.

Figure 2.

(A) Decay-corrected whole-body coronal microPET images of athymic female nude mice bearing U87MG tumor at 5, 15, 30, 60, 120, and 180 min after injection of ¹⁸F-FPRGD4 (3.7 MBq [100 μCi]). (B) Decay-corrected whole-body coronal microPET images of c-neu oncomice at 30, 60, and 150min (5-min static image) after intravenous injection of ¹⁸F-FPRGD4. (C) Decay-corrected whole-body coronal microPET images of orthotopic MDA-MB-435 tumor-bearing mouse at 30, 60, and 150 min after intravenous injection of ¹⁸F-FPRGD4. (D) Decay-corrected whole-body coronal microPET images of DU-145 tumor-bearing mouse (5-min static image) after intravenous injection of ¹⁸F-FPRGD4. (E) Coronal microPET images of a U87MG tumor-bearing mouse at 30 and 60 min after coinjection of ¹⁸F-FPRGD4 and a blocking dose of c(RGDyK). Arrows indicate tumors in all cases. Reproduced from Wu et al. ¹⁴⁹ with permission.

Besides RGD peptides, in vivo imaging using Abegrin, a humanized monoclonal antibody against human integrin α_vβ₃, has been performed after DOTA conjugation and ⁶⁴Cu labeling. MicroPET studies revealed that ⁶⁴Cu-DOTA-Abegrin had a high tumor activity accumulation up to 49.41 ± 4.54% injected dose/g at 71-hour postinjection for U87MG tumors ¹⁵². Not only malignant diseases, the integrin expression after myocardial infarction has also been monitored with ¹⁸F-Galacto-RGD in a Wister rat model. PET imaging and autoradiography revealed focal accumulation in the infarct area started at day 3, peaked between 1 and 3 weeks, and decreased to day 3 level at 6 months after reperfusion. The time course of focal tracer uptake paralleled vascular density as measured by CD31 immunohistochemical analysis, indicating that ¹⁸F-Galacto-RGD is promising for the monitoring of myocardial repair processes ¹⁵³. The results from this study encourage the application of RGD PET imaging to monitor the angiogenesis in other non-cancer diseases.

PET Imaging of VEGF and Its Receptors

VEGF, a potent mitogen in embryonic and somatic angiogenesis, plays a pivotal role in both normal vascular tissue development and many disease processes^{154, 155}. The VEGF family is composed of seven members with a common VEGF homology domain: VEGF-A, VEGF-B, VEGF-C, VEGF-D, VEGF-E, VEGF-F, and placenta growth factor⁵. VEGF-A is a dimeric, disulfide-bound glycoprotein existing in at least seven homodimeric isoforms, consisting of 121, 145, 148, 165, 183, 189, or 206 amino acids. Besides the difference in molecular weight, these isoforms also differ in their biological properties such as the ability to bind to cell surface heparin sulfate proteoglycans⁵.

The angiogenic actions of VEGF are mainly mediated via two endothelium-specific receptor tyrosine kinases, Flt-1 (VEGFR-1) and Flk-1/KDR (VEGFR-2)¹⁵⁶. Both VEGFRs are largely restricted to vascular endothelial cells and all VEGF-A isoforms bind to both VEGFR-1 and VEGFR-2. It is now generally accepted that VEGFR-1 is critical for physiologic and developmental angiogenesis and its function varies with the stages of development, the states of physiologic and pathologic conditions, and the cell types in which it is expressed^{5, 155}. VEGFR-2 is the major mediator of the mitogenic, angiogenic, and permeability-enhancing effects of VEGF. Over-expression of VEGF and/or VEGFRs has been implicated as poor prognostic markers in various clinical studies⁵. Agents that prevent VEGF-A binding to its receptors¹⁵⁷, antibodies that directly block VEGFR-2^{158, 159}, and small molecules that inhibit the kinase activity of VEGFR-2 thereby block growth factor signaling^{65, 160, 161}, are all currently under active development.

The critical role of VEGF-A in cancer progression has been highlighted by the approval of the humanized anti-VEGF monoclonal antibody bevacizumab (Avastin; Genentech) for first line cancer treatment¹⁶². Development of VEGF- or VEGFR-targeted molecular imaging probes could serve as a new paradigm for the assessment of anti-angiogenic therapeutics and for better understanding the role and expression profile of VEGF/VEGFR in many angiogenesis-related diseases. VEGF/VEGFR has been imaged by various imaging modalities though PET is the dominant technique for direct VEGF/VEGFR imaging¹⁶³. In the clinical setting, the right timing can be critical for VEGFR-targeted cancer therapy and non-invasive imaging of VEGF/VEGFR can help in determining whether to start and when to start VEGFR-targeted treatment. With the development of new tracers with better targeting efficacy and desirable pharmacokinetics, clinical translation will be critical for the maximum benefit of VEGF-based imaging agents.

VEGF imaging has been investigated especially with radiolabeled specific antibodies¹⁶⁴. VG76e, an IgG1 monoclonal antibody that binds to human VEGF, was labeled with ¹²⁴I for PET imaging of solid tumor xenografts in immune-deficient mice¹⁶⁵. Whole-animal PET imaging studies revealed a high tumor-to-background contrast. Although VEGF specificity in vivo was demonstrated in this report, the poor immunoreactivity (< 35%) of the radiolabeled antibody limits the potential use of this tracer. HuMV833, the humanized version of a mouse monoclonal anti-VEGF antibody MV833, was also labeled with ¹²⁴I and the distribution and biological effects of HuMV833 in patients in a phase I clinical trial were investigated ¹⁶⁶. Patients with progressive solid tumors were treated with various doses of HuMV833 and PET imaging using ¹²⁴I-HuMV833 was carried out to measure the antibody distribution in and clearance from tissues. It was found that antibody distribution and clearance were quite heterogeneous not only between and within patients but also between and within individual tumors. Bevacizumab, a humanized monoclonal antibody against VEGF, has been labeled with ¹¹¹In to image VEGF-A expression in nude mice model or patients with colorectal liver metastases¹⁶⁷. Although enhanced uptake of ¹¹¹In-bevacizumab in the liver metastases was observed in 9 of the 12 patients, there was no correlation between the level of ¹¹¹In-antibody accumulation and the level of VEGF-A expression in the tissue as determined by in situ hybridization and ELISA ¹⁶⁷. Bevacizumab has also been labeled with the PET isotope ⁸⁹Zr for noninvasive in vivo VEGF visualization and quantification. On small-animal PET images, radiolabeled bevacizumab showed higher uptake compared with radiolabeled human IgG in a human SKOV-3 ovarian tumor xenograft. Tracer uptake in other organs was seen primarily in the liver and spleen (Figure 3)¹⁶⁴. A recent study showed that there was a significant ¹⁸F-FDG kinetics correlation between k1 (The transport coefficient ) and VEGF-A mRNA level determined by gene chip assay (r = 0.51), indicating the possibility to predict the gene expression of VEGF-A with the regression functions from the FDG PET parameters¹⁶⁸.

FIGURE 3.

(A) Coronal CT image and fusion of microPET and CT images (168 h after injection) enables adequate quantitative measurement of ⁸⁹Zr-bevacizumab in the tumor; (B) Coronal planes of microPET images after injection of ⁸⁹Zr-bevacizumab. Reproduced from Nagengast et al. ¹⁶⁴ with permission.

VEGF/VEGFR interactions is one of the most extensively studied angiogenesis-related signaling pathways ¹⁶³. The alternative to overcome the difficulty induced by the soluble and more dynamic nature of VEGF is to image VEGFRs, another indicators of angiogenesis which have superior accessibility. VEGF isoforms exist in nature and have very strong binding affinity and specificity to VEGFRs^{163, 169}. Therefore, a generic strategy is to label these VEGF isoforms with radionuclides to image VEGFRs expression. VEGF₁₂₁ is a soluble, non-heparin-binding variant that exists in solution as a disulfide-linked homodimer containing the full biological and receptor-binding activity of the larger variants⁵. VEGF₁₂₁ has been labeled with ⁶⁴Cu (t_1/2 = 12.7 h) for PET imaging of tumor angiogenesis and VEGFR expression¹⁷⁰. DOTA-VEGF₁₂₁ exhibited nanomolar receptor binding affinity (comparable to VEGF₁₂₁) in vitro. MicroPET imaging revealed rapid, specific, and prominent uptake of ⁶⁴Cu-DOTA-VEGF₁₂₁ (10~15%ID/g) in highly vascularized small U87MG tumor (60 mm³ ) with high VEGFR-2 expression but significantly lower and sporadic uptake (~ 3%ID/g) in large U87MG tumor (1,200 mm³ ) with low VEGFR-2 expression (Figure 4). Western blotting of tumor tissue lysate, immunofluorescence staining, and blocking studies with unlabeled VEGF₁₂₁ confirmed that the tumor uptake is VEGFR specific. This was the first report on PET imaging of VEGFR expression. This study also demonstrated the dynamic nature of VEGFR expression during tumor progression in that even for the same tumor model, VEGFR expression level can be dramatically different at different stages. Successful demonstration of the ability of ⁶⁴Cu-DOTA-VEGF₁₂₁ to visualize VEGFR expression in vivo should allow for clinical translation of this tracer to image tumor angiogenesis and to guide VEGFR-targeted cancer therapy ¹⁷⁰. Further studies showed that the uptake of ⁶⁴Cu-DOTA-VEGF₁₂₁ in the tumor peaked when the tumor size was about 100–250 mm³. Both small and large tumors had lower tracer uptake indicating a narrow range of tumor size with high VEGFR-2 expression¹⁷¹. In another follow-up study, a VEGFR specific fusion toxin VEGF₁₂₁/rGel (composed of VEGF₁₂₁ linked with a G₄S tether to recombinant plant toxin gelonin) was used to treat orthotopic glioblastoma in a mouse model¹⁶⁹. Before initiation of treatment, microPET imaging with ⁶⁴Cu-labeled VEGF₁₂₁/rGel was performed to evaluate the tumor targeting efficacy and the pharmacokinetics. It was found that ⁶⁴Cu-DOTA-VEGF₁₂₁/rGel exhibited high tumor accumulation/retention and high tumor-to-background contrast up to 48 h after injection in glioblastoma xenografts. Based on the in vivo pharmacokinetics of ⁶⁴Cu-DOTA-VEGF₁₂₁/rGel, VEGF₁₂₁/rGel was administered every other day for the treatment of orthotopic U87MG glioblastomas. Histologic analysis revealed specific tumor neovasculature damage after treatment with 4 doses of VEGF₁₂₁/rGel ¹⁶⁹. ⁶⁴Cu was also used to site-specifically label VEGF₁₂₁ and it was found that PEGylation showed considerably prolonged blood clearance. Compared with ^99mTc-labeled analog where the tumor uptake (~ 2 %ID/g) was lower than most of the normal organs and the kidney uptake was about 120 %ID/g, the PEGylated version gave higher tumor uptake (~ 2.5 %ID/g) and lower kidney uptake at about 65 %ID/g ¹⁷².

Figure 4.

MicroPET of ⁶⁴Cu-DOTA-VEGF₁₂₁ in U87MG tumor-bearing mice. (A) Serial microPET scans of large and small U87MG tumor-bearing mice injected intravenously with 5–10 MBq of ⁶⁴Cu-DOTA-VEGF₁₂₁. Mice injected with ⁶⁴Cu-DOTA-VEGF₁₂₁ 30 min after injection of 100 μg VEGF₁₂₁ are also shown (denoted as “Small tumor + block”). (B) Two-dimensional whole-body projection of the 3 mice shown in A at 16 h after injection of ⁶⁴Cu-DOTA-VEGF₁₂₁. Tumors are indicated by arrows. Reproduced from Cai et al. ¹⁷⁰ with permission.

PET imaging using radiolabeled VEGF can also play a role in other angiogenesis-related diseases besides cancer. Myocardial infarction (MI) can lead to the activation of many biological pathways including VEGF/VEGFR signaling^{173, 174}. Using the previously validated PET tracer ⁶⁴Cu-DOTA-VEGF₁₂₁, the kinetics of VEGFR expression was imaged for the first time in living subjects using a rat model of MI¹⁷⁵. MI was induced by ligation of the left anterior descending coronary artery in Sprague-Dawley rats and confirmed by ultrasound. ⁶⁴Cu-DOTA-VEGF₁₂₁ PET scans were performed prior to MI induction, and at days 3, 10, 17, and 24 after MI induction. Baseline myocardial uptake of ⁶⁴Cu-DOTA-VEGF₁₂₁ was minimal (0.3 ± 0.1 %ID/g). After MI, ⁶⁴Cu-DOTA-VEGF₁₂₁ myocardial uptake significantly increased (up to 1.0 ± 0.1 %ID/g) and was elevated for 2 weeks, after which it returned to baseline levels. In a hindlimb ischemia model, PET imaging showed significantly higher ⁶⁴Cu-DOTA-VEGF₁₂₁ uptake in ischemic hindlimbs than in non-ischemic hindlimbs. Treadmill exercise training was also found to increase ⁶⁴Cu-DOTA-VEGF₁₂₁ uptake in ischemic hindlimbs compared with non-exercised hindlimbs¹⁷⁶. With ⁶⁴Cu-DOTA-VEGF₁₂₁ PET imaging, we have evaluated the VEGFR expression kinetics noninvasively in a rat stroke model. The results revealed that the tracer uptake in the stroke border zone peaked at approximately 10 days after surgery, indicating neovascularization as confirmed by histology (VEGFR-2, BrdU, and lectin staining)¹⁷⁷.

All VEGF-A isoforms bind to both VEGFR-1 and VEGFR-2⁵. In the imaging studies reported to date, specificity to either VEGFR-1 or VEGFR-2 has rarely been achieved as most of the tracers are based on VEGF isoforms. Kidneys have high VEGFR-1 expression which can take up VEGF-A based tracer thus usually makes it the dose limiting organ^{170, 178}. Alanine-scanning mutagenesis has been used to identify a positively charged surface in VEGF₁₆₅ that mediates the binding to VEGFR-2¹⁷⁹. Arg⁸², Lys⁸⁴, and His⁸⁶, located in a hairpin loop, were found to be critical for binding VEGFR-2, while negatively charged residues, Asp⁶³, Glu⁶⁴, and Glu⁶⁷, were associated with VEGFR-1 binding. Mutations in the 63–67 region of VEGF exhibited only modest effects on VEGFR-2 binding but significant reduction in affinity with VEGFR-1. Recently, our lab engineered a D63AE64AE67A mutant of VEGF₁₂₁ (VEGF_DEE) by recombinant DNA technology to develop a VEGFR-2-specific PET tracer. Cell binding assay demonstrated that VEGF_DEE had about 20-fold lower VEGFR-1 binding affinity and only slightly lower VEGFR-2 binding affinity as compared with VEGF₁₂₁. Both ⁶⁴Cu-DOTA-VEGF₁₂₁ and ⁶⁴Cu-DOTA-VEGF_DEE had rapid and prominent activity accumulation in VEGFR-2-expressing 4T1 tumors. However, the renal uptake of ⁶⁴Cu-DOTA-VEGF_DEE was significantly lower than that of ⁶⁴Cu-DOTA-VEGF₁₂₁ as rodent kidneys expressed high levels of VEGFR-1, indicating that VEGF_DEE is superior to wild type VEGF₁₂₁ for imaging tumor angiogenesis (Figure 3) ¹⁸⁰. The DOTA conjugation of VEGF proteins in this study is randomly instead of site-specifically. It will be critical to further develop more potent VEGFR-2 specific mutants, site-specifically label VEGF analog proteins with various isotopes including ⁶⁴Cu to improve angiogenesis imaging quality and result analysis ¹⁷⁵.

Imaging of Matrix Metalloproteinases

Matrix metalloproteinases (MMPs) are a family of zinc- and calcium-dependent endopeptidases which are responsible for the enzymatic degradation of connective tissue und thus facilitate endothelial cell migration during angiogenesis ¹⁸¹. Additionally, MMPs process and release bioactive molecules such as growth factors, proteinase inhibitors, cytokines and chemokines¹⁸². From the more than 18 members of the MMP family, the gelatinases MMP-2 and -9 are most consistently detected in malignancies¹¹⁷. In the progression of the atherosclerotic lesions, MMP-3 and -9 have been shown to limit plaque growth and promote a stable plaque phenotype and MMP-12 supports atherosclerotic lesion expansion and destabilization ¹⁸³. Many strategies have been developed to image MMPs level for the assessment of angiogenesis ^{184, 185}.

The so-called “smart probes” have been developed to contain fluorescent dyes and MMP cleavable sequences ^{186, 187}. It has been reported a MMP-2-sensitive probe was activated by MMP-2 in vitro, producing up to an 850% increase in near-infrared fluorescent signal intensity, and MMP-2-positive tumors were easily identified as high-signal-intensity regions as early as 1 hour after intravenous injection of the MMP-2 probe ¹⁸⁸.

Via phage display techniques, the MMP specific decapeptide H-Cys-Thr-Thr-His-Trp-Gly-Phe-Thr-leu-Cys-OH (CTT) was found and could be labeled with ¹²⁵I and ^99mTc. However, this tracer has unfavourable characteristics for in vivo imaging because the metabolic stability of the compound is low and lipophilicity is high ¹⁸⁹. Another group labeled this peptide with ¹¹¹In after conjugating it with a highly hydrophilic and negatively charged chelator DTPA. A significant correlation was observed between the accumulation in the tumor as well as tumor-to-blood ratio of ¹¹¹In-DTPA-CTT and gelatinase activity. Moreover, ¹¹¹In-DTPA-CTT showed low levels of radioactivity in the liver and kidneys¹⁹⁰. CTT peptide also has been labeled with ⁶⁴Cu after DOTA conjugation for PET imaging of MMP. ⁶⁴Cu-DOTA-CTT inhibited hMMP-2 and mMMP-9 with similar affinity to CTT. MicroPET imaging studies showed that ⁶⁴Cu-DOTA-CTT was taken up by MMP-2/9-positive B16F10 murine melanoma tumors, however, the low affinity for MMP-2 and MMP-9 and in vivo instability of CTT-based imaging probes need to be overcome for further applications¹⁹¹.

Another approach is to label small molecule MMP inhibitors (MMPIs), which are typically used as antiangiogenic drugs. In general, MMPIs possess a zinc binding group (ZBG) complexing the zinc ion of the active site and are classified into several groups owing to their lead structures¹⁸¹. Different ¹⁸F and ¹¹C labeled MMPIs have been synthesized and evaluated preclinically with mixed results^{192, 193}. Fluorinated MMPIs based on lead structures of the broad-spectrum inhibitors N-hydroxy-2(R)-[[(4-methoxyphenyl)sulfonyl](benzyl)-amino]-3-methyl-butanamide (CGS 25966) and N-hydroxy-2(R)-[[(4-methoxyphenyl)sulfonyl](3-picolyl)-amino]-3-methyl-butanamide (CGS 27023A) have been synthesized and showed high in vitro MMP inhibition potencies for MMP-2, MMP-8, MMP-9, and MMP-13 ¹⁹⁴. However, in vivo microPET study with ¹¹C-CGS 25966 failed to demarcate MMP positive tumors¹⁹⁵. A ¹¹C-labeled MMPI (2R)-2-[[4-(6-fluorohex-1-ynyl)phenyl]sulfonylamino]-3-methylbutyric acid ¹¹C-methyl ester (¹¹C-FMAME), has also been synthesized and applied to two animal models of breast cancer, MCF-7 xenograft transfected with IL-1 and MDA-MB-435 xenograft in athymic mice. Again, low tumor-to-blood and tumor-to muscle ratios of these tracers do not allow visualization of the tumors in microPET studies^{192 196}. However, biodistribution study with ¹⁸F-labeled similar compound, (2R)-2-[4-(6-¹⁸F-Fluorohex-1-ynyl)-benzenesulfonylamino]-3-methylbutyric acid (¹⁸F-SAV03), showed higher tumor uptake of the tracer than normal organs¹⁹³. Other MMPIs have also been synthesized and labeled with radionuclides including ¹¹¹In and ¹⁸F ^197–194. Nevertheless, significant improvements in tumor MMP targeting and in vivo pharmacokinetics are necessary before the use of MMP radiotracer imaging will be translated into the clinic.

Imaging of other angiogenesis related targets

Fibronectin is a large glycoprotein, which can be found physiologically in plasma and tissues. However, the extra-domain B of fibronectin (EDB), consisting of 91 amino acids, is not present in the fibronectin molecule under normal conditions, except for the endometrium in the proliferative phase and some vessels of the ovaries. EDB is interesting as a marker of angiogenesis as it is expressed in a variety of solid tumors, as well as in ocular angiogenesis and wound healing¹⁹⁸. The human antibody fragment scFv(L19) has been shown to efficiently localize on neovasculature both in animal models and in cancer patients. In a study with patients suffering from various solid tumors, 16 of 20 tumor lesions could be identified by SPECT using ¹²³I-scFv(L19). Whether the unidentified tumors were not detected because they were either in a phase of slow growth with low levels of angiogenesis, or due to the technical limitations of SPECT imaging, is not clear¹⁹⁹. No reports about PET tracers targeting EDB are available up to now. Other angiogenesis related biomarkers such as angiopoietins/Tie receptors²⁰⁰ and CD276²⁰¹ are also potential targets for angiogenesis imaging. Angiopoietins/Tie receptors are involved in regulation of complex interactions between endothelium and surrounding cells. CD276 has been observed to be overexpressed in tumor versus normal endothelium.

Summary and perspective

Numerous imaging techniques are available for assessing tissue vasculature on a structural, functional and molecular level. A wide variety of targeting ligands (small molecules, peptides, peptidomimetics, and antibodies) have been conjugated with various imaging labels for MRI, ultrasound, optical, SPECT, PET, and multimodality imaging of angiogenesis. All these methods have been successfully used preclinically and will hopefully aid in antiangiogenic drug evaluation in animal studies. Due to its high sensitivity and low amounts of tracer that have to be used, PET will probably be the first to be used on a wide scale in patients in the intermediate term. In addition, toxicity issues of PET tracers are of less importance compared to MRI or Ultrasound imaging probes since only pica molar amount will be used for imaging purpose. However, it is likely that not one single parameter, target structure or imaging technique will be used for the assessment of angiogenesis in the future, but rather a multimodality, multiplexing imaging which will allow for evaluation of the angiogenic cascade in its full complexity to acquire comprehensive information. It is predictable that the new generation clinical PET/CT and microPET/microCT, as well as PET/MRI and microPET/microMRI currently in active development ^202–204, will likely play a major role in molecular imaging of angiogenesis for the years to come.

Although it is generally assumed that non-invasive imaging results correlate with the target expression level, such assumption has not been extensively validated. In most reports, two tumor models are studied where one acts as a positive control and the other as a negative control. Quantitative correlation between the target expression level in vivo and the non-invasive imaging data is rare^205–208. Such correlation is critical for future therapeutic response monitoring, as it would be ideal to be able to monitor the changes in the target expression level quantitatively, rather than qualitatively, in each individual patient. Lack of accurate quantification is one of the hindrances why only a few radiotracers including PET tracers have been used in humans up to now, and their role in assessment of anti- or pro-angiogenic therapies is still unsettled.

To further improve imaging of the angiogenesis process at molecular level, it is necessary to identify new angiogenesis related targets and corresponding specific ligands and to optimize currently available imaging probes. Thorough and full understanding of the physiological and pathological changes during angiogenesis will be critical for new targets identification. Optimization of current available imaging probes can be achieved in several ways. First, oligomerization (homo or hetero) the targeting ligand (typically peptide) can improve the binding affinity as well as tissue retention likely due to the polyvalency effect ²⁰⁹. Second, site-specific labeling may be advantageous than randomly labeling on lysine residues in terms of retaining the binding affinity and functional activity ¹⁷⁵. Third, incorporation of a linker between the targeting ligand and the label can improve the pharmacokinetic properties. Glycosylation, PEGylation, and various other linkers have been shown to improve the imaging quality. Last, development of new strategies to improve the labeling yield (most applicable to ¹⁸F-based tracers) is critical for future clinical studies. To foster the continued discovery and development of angiogenesis-targeted imaging agents, cooperative efforts are needed from cellular/molecular biologists to identify and validate novel imaging targets, chemists/radiochemists to synthesize and characterize the imaging probes, and engineers/medical physicists/mathematicians to develop high sensitivity/high resolution imaging devices/hybrid instruments and better image reconstruction algorithms.

Non-invasive imaging of angiogenesis has clinical applications in many aspects including lesion detection, patient stratification, new drug development/validation, treatment monitoring, and dose optimization. For example, glucosamino ^99mTc-d-c(RGDfK) gamma-camera imaging has been applied to monitor the therapeutic efficacy of paclitaxel in LLC tumor-bearing mice ²¹⁰. With the development of new tracers with better targeting efficacy and desirable pharmacokinetics, clinical translation will be critical for the maximum benefit of these imaging probes. Most of the molecular imaging probes suffer from the slow translation from bench to bedside. Multiple steps in pre-clinical development, especially the investigational new drug (IND)-directed toxicology, significantly slowed down the process of converting a newly developed agent into a diagnostic imaging probe for clinical testing. The high specificity required for molecular imaging not only leads to higher costs of development but also smaller market potential, which may make them considered too risky by investors for commercial development. However, the situation has gradually changed over the last several years thanks to the continued development and wider availability of scanners dedicated to small animal imaging studies, as well as the exploratory IND mechanism proposed by the food and drug administration (FDA) to allow faster first-in-human studies. Now the molecular imaging techniques can bridge the gap between preclinical and clinical research to develop candidate drugs that have the optimal target specificity, pharmacodynamics, and efficacy. It is expected that in the foreseeable future, anigogenesis imaging with PET tracers will be routinely applied in anti-cancer clinical trials, paving the way to personalized molecular therapy.