Background
Positron emission tomography (PET) is one of the fastest growing medical imaging modalities worldwide and, alone or in combination with other non-invasive imaging techniques, is used as a scientific and diagnostic tool in many medical fields [
1].
A variety of radiopharmaceuticals has been established for imaging with PET, allowing visualization, monitoring, and measuring molecular and cellular events in the living organisms with high sensitivity (reviewed in [
2]).
Angiogenesis is driven by potent pro-angiogenic factors and signaling molecules, including growth factors and growth factor receptors [
3]. One of the most prominent of these is vascular endothelial growth factor (VEGF)A, also known as vascular permeability factor [
4]. Although VEGFA binds with both VEGFR-1 and VEGFR-2 receptors, it is commonly agreed that VEGFR-2 is the key mediator of the mitogenic, angiogenic, and microvascular permeability-enhancing effects of VEGF [
5]. The overexpression of VEGF/VEGFR-2 by tumor endothelium is associated with increased angiogenesis, metastatic spread of tumor cells, and with poor prognosis in cancer patients [
6].
A review of the current literature reveals ten anti-VEGF/VEGFR-2 drugs, approved by the US Food and Drug Administration (FDA), in clinical use as monotherapy or in combination for the treatment of various types of cancer. These drugs include antibodies and their fragments (Fab-fragments, single chains), proteins, peptides, and tyrosine kinase inhibitors (TKIs) [
7,
8]. Bevacizumab (BVZ, Avastin®), a humanized anti-VEGF monoclonal antibody, was the first anti-angiogenic drug approved by the FDA in 2004–2006 for the treatment, in combination with chemotherapy, of patients with metastatic colorectal cancer, advanced non-small cell lung cancer, and renal cell carcinoma (reviewed in [
9]). The European Medicines Agency retains BVZ in combination with paclitaxel (PTX) or capecitabine as the first-line treatment of patients with HER-2-negative locally recurrent/metastatic breast cancer [
10].
The types of drugs mentioned above as well as several other molecules have demonstrated potential as PET ligands for non-invasive in vivo imaging of the VEGF/VEGFRs. In preclinical studies, several zirconium-89 [
11,
12] and copper-64 [
13] labeled antibodies showed promising ability to visualize and quantify VEGF/VEGFR levels in tumor vasculature. In clinical applications, PET with zirconium-89-labeled BVZ has been used for assessment of anti-angiogenic treatment efficacy in patients with metastatic renal cell carcinoma [
14] and non-small cell lung cancer patients [
15].
Most TKIs are multi-targeting agents. There is therefore a growing interest in the discovery of TKIs with improved target selectivity, affinity in the subnanomolar range, and capability to penetrate the cell membrane [
16]. A number of TKIs that block the adenosine triphosphate binding site of the VEGFRs TK domain and inhibit receptor-mediated intracellular signaling, thereby reducing angiogenesis, are among the new candidates. One of these, vandetanib (ZD6474) is an orally active VEGFR-2 TKI that has been shown to suppress tumor-induced angiogenesis in several xenograft models [
17].
3-Piperidinylethoxy-anilinoquinazoline (PAQ) is an analog to vandetanib with 40 times stronger inhibitory properties for the VEGFR-2 [
18]. The (
R)-PAQ molecule has two stereoisomers,
S and
R, with IC
50-values of 10 and 1 nM, respectively, for the VEGFR-2 at competitive concentrations of 2 μM adenosine triphosphate. Regarding the specific binding in comparison to others RTKs, the VEGFR-2
R-isomer had a 200-fold higher affinity versus epidermal growth factor receptor (EGFR) compare to only a 10-fold difference for
S-isomer versus EGFR. This data convinced us to perform the further studies with the pure
R-isomer.
We have previously described the synthesis and carbon-11 labeling of PAQ to yield (
R)-[
11C]PAQ [
19] and demonstrated that the radiotracer uptake correlated with high VEGFR-2 expression in primary tumors and during metastasis development [
20].
The current study aimed to examine the capability of using the (
R)-[
11C]PAQ VEGFR-2-targeting for monitoring anticancer treatment in the MMTV-PyMT/FVB (PyMT) transgenic mouse breast cancer model. This animal model was chosen for its translational capacity, i.e., developing adenocarcinomas with metastatic potential and its similarities to human luminal B breast tumors [
21]. The PTX/BVZ therapies and the dosing selected for this study were based on our previous pilot studies and other preclinical studies [
22,
23]. PTX is a mitotic inhibitor commonly used as a first-line chemotherapy. When combined with BVZ, progression-free survival and objective response rate in patients with metastatic breast cancer were significantly improved compared to PTX alone [
24,
25]. Since BVZ has a high specificity for only human VEGFA, its murine analog B20-4.1.1 was used in the present study [
26].
Discussion
This study investigated the potential capability of using the VEGFR-2-targeting (
R)-[
11C]PAQ PET radiotracer to monitor and evaluate the efficacy of anticancer treatment in the PyMT mouse model of breast cancer. The study was performed in female mice at the late stage of malignancy, which is characterized by high expression levels of VEGFR-2, CD31, and other proangiogenic factors in the mammary tumor vasculature [
29]. Histological profiling of the PyMT tumors, also included in this study, showed typically heterogeneous histology patterns, with irregularly distributed necrosis and more clustered areas with higher mitotic index and angiogenic activity. These factors also contribute to the heterogeneous distribution of the VEGFR-2-targeting (
R)-[
11C]PAQ radiotracer within the tumors observed here.
The quantitative analysis of the (R)-[11C]PAQ PET data revealed a significant reduction of the radiotracer uptake (SUVmax) in the PyMT mammary tumors within both the B20-4.1.1/PTX combination and B20-4.1.1 monotherapy treatment groups compared to the control (VEH group). The result was statistically significant in these two groups when analyzed both within the group and in comparison to the control (VEH group).
Significant reductions of the mammary TVs during therapy, as measured with MR imaging, were observed only within the B20-4.1.1/PTX combination treatment group, and only the modest effects were observed in mice treated with PTX and B20-4.1.1 monotherapies. It is important to note that, even though TV was not significantly reduced in the B20-4.1.1 group, the TV reduction was higher in this group than in the VEH group. We detected a large variation in relative TV changes within the B20-4.1.1 group, even though there was a consistent reduction in the SUVmax from day 0 to day 4. We interpret this as an effect of anti-angiogenic treatment with B20-4.1.1 on the tumor microvasculature leading to a decreased uptake of the targeting radiotracer, but not to a reduction in tumor size. Although it appears that PTX, which is not as closely associated with angiogenesis, does not lead to a systematic reduction in SUVmax (compared to VEH), a corresponding lack of effect of PTX on tumor volume or number of Ki67 positive cells prevents us from presenting this as evidence of tracer selectivity for tumor angiogenic sites.
We observed a trend toward a positive correlation between the (R)-[11C]PAQ SUVmax changes and the mammary TV reductions in the B20-4.1.1/PTX group. The moderate (rs = 0.45) correlation could possibly be due to the large spread in initial tumor sizes (and therefore their baseline characteristics) in the treatment group. In this model and study protocol, the (R)-[11C]PAQ SUVmax appeared to be a more sensitive to treatment than SUVmean. The SUVmean values were found to be more variable due to high diversity in tumors histological pattern and intra-tumoral heterogeneity.
The current study showed that the microvascular density of the PyMT tumors was significantly lower in the B20-4.1.1/PTX combination treatment group than in the other groups. A number of preclinical studies have demonstrated that anti-angiogenic drugs enhance chemotherapy delivery and penetration, improving tumor response by remodeling the vasculature [
30,
31]. Dickson et al. showed that a single dose of the anti-VEGF antibody BVZ caused an overall decrease in tumor microvascular density by destroying the immature vessels and improving tumor perfusion and responsiveness to chemotherapy in neuroblastoma xenografts [
32].
PAQ acts as a competitive inhibitor of the ATP-binding pocket at the catalytic intracellular tyrosine-kinase (TK) domain of VEGFR-2. Activation of VEGF-2 by VEGF results in the formation of receptor dimers, following by cross-phosphorylation of the intracellular TK domains of the receptors and intracellular signal transduction [
33].
PAQ binds to the TK domain only when the receptor is in its inactive conformation, at which time the ATP pocket is available (i.e., in the absence of ligand binding/dimerization/phosphorylation). Thus, the balance between all the factors that affect the availability of the ATP-binding domain at a given time will determine the amounts of radiolabeled PAQ retained at that imaging session. The production and release of VEGF are higher when the tumors are fast growing and hypoxic [
34], as in the PyMT model. The higher the levels of VEGF, the greater the probability that it will interact with the receptor and the lower the number of “free” ATP-binding sites. Activation by VEGF results in receptor internalization, endocytosis, and recycling, but the VEGFR-2 undergoes constitutive endosome-to-plasma membrane recycling even in the absence of ligand [
35]. The dynamics of this recycling will affect the speed at which the ATP-binding sites once again become available for PAQ binding. VEGF-targeted therapies like BVZ would initially lead to a decrease in the VEGF available for binding with the receptor. Therefore, in the acute phase, the relative availability of the ATP-binding sites for (
R)-[
11C]PAQ could increase. However, with time, the tumor endothelial cells die and the blood vessel regression is achieved (reviewed in [
36]), which would lead to a decreased retention of (
R)-[
11C]PAQ. Dynamic changes in the concentration of circulating VEGF and the contribution of host stromal VEGF make it difficult to estimate the amount of antibodies for efficient blocking [
26].
B20-4.1.1 is a cross-species monoclonal antibody targeting both human and murine VEGF [
26], and it has been used to treat various preclinical tumor models [
37,
38]. Anti-VEGF blocking depends on both the tumor context and treatment. Bagri et al. [
39] have evaluated the effects of anti-VEGF treatment in a diverse panel of tumor xenografts and genetic mouse models of cancer. Their studies concluded that continuous VEGF suppression with B20-4.1.1 provided additional benefit in reducing tumor growth when combined with chemotherapy. However, there have been only a few reports on the use of B20-4.1.1 in the MMTV-PyMT model. A recently published study [
40] demonstrated that long-term monotherapy with B20-4.1.1 caused significant tumor growth inhibition in the PyMT model and affected microvessel density in a similar manner as the two anti-angiogenic TKIs, nintedanib, and dovitinib.
In contrast to B20-4.1.1 monotherapy and B20-4.1.1/PTX combination therapy, no significant treatment-induced changes in TV and radiotracer uptake were observed in the PTX monotherapy group. PTX stabilizes the microtubules in proliferating cells by blocking them from the progression of mitosis, and it induces apoptosis in cancer cells [
41,
42]. Recent studies have, however, demonstrated that PTX induced resistance to chemotherapy and promoted pulmonary and lymphatic metastasis in the PyMT model. Volk-Drapper et al. [
43] have shown that repeated PTX treatment caused pro-oncogenic and intratumoral inflammatory changes in the PyMT mammary tumors through activation of the Toll-like receptor (TLR4). Another study [
44] showed that high-dose PTX treatment in PyMT mice caused increased macrophage infiltration that protected tumors from cell death and facilitated tumor progression and metastasis.
In our study, possible PTX effects on proliferation were examined by immunohistochemical determination of the proliferation marker Ki67. Ki67 has been identified as an independent prognostic factor in breast cancer patients [
45] and has also been used to evaluate PyMT tumor proliferating activity in preclinical studies [
46]. In our study, a single dose of PTX did not alter the fraction of Ki67 positive cells in any treatment group. Only the tumors treated with combined B20-4.1.1/PTX showed clearly reduced Ki67 proliferation index on day 4 compared to other groups, though this difference did not reach statistical significance.
The structural analogs of (
R)-[
11C]PAQ, carbon-11-labeled vandetanib, and chloro-vandetanib have been successfully developed for potential applications as VEGFR-2 radiotracers; however, they have yet to be evaluated in vivo [
47]. In this study, we used the PAQ synthesis protocols that we have used in previous validations of this radiotracer [
18]. General radiolabeling and purification procedures with carbon-11 by alkylation reactions with [
11C]methyl iodide are well-established methods. The encouraging results obtained during the [
11C]PAQ evaluation in vitro and in vivo motivated us to here further evaluate the (
R)-[
11C]PAQ in additional disease models. However, fluorine-18 is an attractive PET radioisotope due to its longer half-life (permitting, for example, multiple studies from the same batch and higher imaging resolution), and future comparative studies with (
R)-[
18F]PAQ could be of interest. Prabhakaran et al. [
48] have developed and synthesized the fluorine-18-labeled fluoroethyl analog of (
R)-[
11C]PAQ, (
R)-[
18F]FEPAQ. The authors have demonstrated tracer’s specific selectivity for VEGFR-2 in human glioblastoma frozen sections, though the tracer has not yet, to our knowledge, been evaluated in vivo.
In the clinical setting, BVZ combined with paclitaxel failed to show an overall survival benefit in metastatic breast cancer patients [
49]. Several mechanisms of resistance to VEGF-targeted therapy have been suggested; among them are a complex interaction between tumor cells and stroma, an increased aggressiveness of the tumor caused by hypoxia (and thus new mutations), hypoxia-induced increase of cancer stem cells, and an activation of alternative pro-angiogenic signaling pathways [
50]. Many other biomarkers for monitoring anti-angiogenic therapy have been studied, including circulating levels of pro-angiogenic factors, mutations in angiogenesis-related genes, tumor microvascular density, levels of vascular perfusion, hypertension, and in situ markers in tumor tissue [
51,
52]. In addition, novel molecular and functional imaging probes targeted angiogenesis have been intensively developed and evaluated (reviewed in [
53]). Despite the encouraging results with some of the above, there is still a lack of biomarkers that can be used to select a population of patients that would benefit from anti-angiogenic therapy.
In our study, we were able to demonstrate the promising capability of (R)-[11C]PAQ PET imaging for visualizing/quantifying treatment response. However, the single dose/short-term treatment was insufficient to produce statistically convincing evidence in the PyMT model. Future studies should examine the ability of (R)-[11C]PAQ to monitor therapeutic response in other dosing protocols. Similarly, multiple sequential (R)-[11C]PAQ PET studies over time could be attempted to see if an even more appropriate time than day 4 for the therapeutic read-out can be found.