This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Moffat, B. A. Articles by Ross, B. D. Search for Related Content PubMed PubMed Citation Articles by Moffat, B. A. Articles by Ross, B. D.

Inhibition of Vascular Endothelial Growth Factor (VEGF)-A Causes a Paradoxical Increase in Tumor Blood Flow and Up-Regulation of VEGF-D

Authors' Affiliations: Departments of ¹ Radiology, ² Radiation Oncology, ³ Biostatistics, and ⁴ Pathology, Center for Molecular Imaging, University of Michigan Medical School, Ann Arbor, Michigan

Requests for reprints: Brian D. Ross, Center for Molecular Imaging, Department of Radiology, University of Michigan, Ann Arbor, MI 48109-0503; E-mail: bdross{at}umich.edu.

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Purpose: Vascular endothelial growth factor (VEGF)-A is an important mediator of angiogenesis in almost all solid tumors. The aim of this study was to evaluate the effect of VEGF-A expression on tumor growth, perfusion, and chemotherapeutic efficacy in orthotopic 9L gliosarcomas.

Experimental Design: Stable 9L cell lines underexpressing and overexpressing VEGF-A were generated. Anatomic, susceptibility contrast, and continuous arterial spin-labeling magnetic resonance imaging were used to quantify the volume, blood volume, and blood flow of tumors orthotopically grown from these and wild-type 9L cells. Histologic, immunohistochemical, and quantitative reverse transcription-PCR analyses were also done on excised tumors. Finally, the effects of carmustine chemotherapy were also evaluated.

Results: Orthotopic tumors underexpressing VEGF-A had slower growth rates (increased median survival), greater blood flow, vessel density, and VEGF-D expression, but no statistical difference in blood volume and chemotherapeutic sensitivity, compared with tumors with wild-type levels of VEGF-A. Tumors overexpressing VEGF-A had faster growth rates, greater blood volume, vessel density, and blood flow but no statistical difference in VEGF-D expression and chemotherapeutic sensitivity compared with wild-type VEGF-A-expressing tumors.

Conclusion: Blood volume and blood flow are independent and different biomarkers of tumor perfusion. Therefore, both should be measured when characterizing the efficacy of antiangiogenic therapies. Underexpression of VEGF-A does not result in complete inhibition of angiogenesis. Moreover, these tumors have a different perfusion phenotype, suggesting that angiogenesis is mediated by an alternative pathway. The results indicate that VEGF-D is a plausible alternative mediator of this angiogenesis.

The main aim of this study was to evaluate magnetic resonance imaging (MRI) biomarkers of blood flow and blood volume to identify differences in three variants of the orthotopic 9L brain tumor model that had different angiogenic potential. Modulation of tumor angiogenesis was achieved by underexpressing (VEGF^–) and overexpressing (VEGF⁺) vascular endothelial growth factor (VEGF)-A with respect to wild-type 9L (VEGF⁰) expression levels. A key mediator of both normal and pathologic angiogenesis, VEGF-A has been implicated in tumor development (8), including rodent (9) and human malignant gliomas (2, 10, 11), through its ability to stimulate angiogenesis by binding to a family of VEGF receptors. Recently, several antiangiogenic tumor therapies have been developed for blocking VEGF-A or its receptors (2). In the current study, the effect of modulating VEGF-A expression levels on MRI measurements of tumor blood flow and blood volume, tumor growth rates, animal survival, tumor histopathology, and chemotherapeutic sensitivity were investigated.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
Cloning of VEGF-A. Total cellular RNA was purified from 9L gliosarcoma cells using the RNeasy mini kit (Qiagen, Valencia, CA). Full-length rat VEGF-A was cloned from rat 9L gliosarcoma cDNA using primers (5'-GCTCTAGAGTCGACGCAATCATGAACTTTCTGCTCTCTTGG-3', 5'-GCTCTAGAGTCGACTCACCGCCTTGGCTTGTCAC-3') based on the rat (glioma derived) VEGF-A cDNA (accession no. NM_031836; ref. 12). The PCR product was cloned into plasmid pEF using the XhoI site. Independent clones containing the entire rat VEGF-A cDNA were then sequenced at the University of Michigan DNA Sequencing Core to confirm the appropriate sequence. Due to the use of a single restriction site on both sides of the cDNA, an antisense VEGF-A plasmid was also selected to make the VEGF^– tumor cell line.

Cell culture conditions and generation of stable cell lines. The parental VEGF⁰ cell line was cultured in DMEM with GlutaMAX (Invitrogen/Life Technologies, Carlsbad, CA), and supplemented with 10% inactivated FCS, penicillin, and streptomycin at 37°C and 5% CO₂. To generate stable cell lines, the corresponding pEF plasmids (VEGF⁺ or VEGF^–) were transfected into the 9L cell line using FuGene6 (Roche, Basel, Switzerland). After 2 weeks of selection in culture medium containing 400 mg/mL G418 (Invitrogen, Carlsbad, CA), single cell-derived stable lines were obtained and tested for secretion of VEGF-A into the culture medium by SDS-PAGE and Western blotting (see below). To characterize growth rates, the three representative 9L cell lines were plated in multiple parallel 60 mm dishes. Four replicate plates were trypsinized and counted using a Coulter Counter every day for 7 consecutive days. In addition, the sulfur rhodamine-B growth inhibition assay was used to evaluate the sensitivity of each cell line to carmustine [1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU); ref. 13].

Intracranial tumor implantation. All of the animal work was carried out in the animal facility at the University of Michigan in accordance with federal, local, and institutional guidelines. Intracerebral brain tumors were implanted in male Fischer 344 rats (Charles River Breeding Laboratories, Wilmington, MA) weighing between 125 and 150 g. The rats were anesthetized by an i.p. administration of a xylazine (13 mg/kg) and ketamine (87 mg/kg) mixture. A small skin incision was made over the right hemisphere. Using a high-speed drill, a 1-mm-diameter burr hole through the skull was created and an inoculum of 1 x 10⁵ VEGF⁰, VEGF⁺, or VEGF^– cells in 5 µL serum-free medium was introduced through a 27-gauge needle to a depth of 3 mm. The rats were allowed to recover after the burr hole was filled with bone wax.

Histopathology and immunohistochemistry. The brains of rats bearing orthotopic tumors were removed and fixed in 10% paraformaldehyde following animal euthanasia. Two days later, the fixed tissues were transferred to 70% ethanol (24 hours) and then embedded in paraffin. Formalin-fixed paraffin-embedded specimens were sectioned at a thickness of 5 µm and mounted on poly-L-lysine–coated slides. After deparaffinization and rehydration, sections were stained with H&E for histopathologic evaluation. For immunohistochemical study, slides obtained from formalin-fixed paraffin-embedded brain specimens were subjected to microwave antigen retrieval, done in citrate buffer (pH 6.0), followed by peroxidase blocking (5 minutes), and incubation in proteinase K solution (DAKO, Glostrup, Denmark). To detect von Willebrand factor and VEGF-D, a rabbit polyclonal antibody for von Willebrand factor (DAKO, Glostrup, Denmark) and a monoclonal antibody for VEGF-D (Abcam, Cambridge, MA), diluted (1:200) with PBS containing 2% goat serum, was applied and incubated for 30 minutes. The slides were washed with PBS, treated with a goat anti-rabbit secondary antibody (DAKO EnVision) detection kit, incubated in a 3,3'-diaminobenzidine chromagen (DAKO) solution, and counterstained with hematoxylin. On the anti–von Willebrand factor slides, 10 hotspots per each tumor type were chosen for evaluation of vascular density (number of vessels per x400 field).

Western blot detection of VEGF proteins. For immunoblot analysis of secreted VEGF-A in cell cultures, all three 9L cell lines were grown to equal density in cell culture medium (see above). After 24 hours, the conditioned medium was removed and filtered through a 0.45 µm filter (Sigma, St. Louis, MO) to remove floating cells and debris, followed by incubation with 50 mL heparin-agarose (Sigma) at 4°C overnight with continuous agitation. Heparin-agarose beads were then spun down and washed thrice with PBS buffer, boiled in reducing Laemmli sample buffer, and resolved by SDS-PAGE. Western analysis was then done using an anti-VEGF-A antibody [Santa Cruz Biotechnology, Inc., Santa Cruz, CA; VEGF(147)]. For immunoblot analysis of VEGF-A and VEGF-D secreted by tumors, orthotopic tumors were harvested from the respective rats, weighed, and frozen immediately in liquid nitrogen. Frozen tumor samples (1 mg tissue per 10 µL) were homogenized using a rotator-stator and 5 mm head (Fischer, Hampton, NH) in radioimmunoprecipitation assay buffer containing protease inhibitors [1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 150 mmol/L NaCl, 50 mmol/L Tris (pH 8.0), 0.2 units/mL aprotinin, 2 µg/mL leupeptin, 1 µg/mL pepstatin A, and 2 mmol/L phenylmethylsulfonyl fluoride, 10 µL per 1 mg tissue]. These were then centrifuged at 1,000 x g for 10 minutes to remove the cell debris. The supernatant was reserved and the concentration of protein estimated using a Bio-Rad detergent-compatible protein assay kit according to the instructions of the manufacturer. Laemmli sample buffer was added to the supernatants to give a final protein concentration of 2 µg/µL. Western analyses were then done using antibodies for VEGF-A (Abcam) and VEGF-D (Abcam). All the immunoblots were visualized using enhanced chemiluminescence (Pierce, Rockford, IL).

Quantification of mRNA for VEGF-D. From the supernatant of homogenized and centrifuged tumor samples (prepared identically to that used for the Western analysis above), RNA was purified using the RNeasy kit in combination with the RNase-Free DNase Set (Qiagen) to remove cellular DNA. The level of mRNA for VEGF-D was determined by quantitative reverse transcription-PCR done on a DNA Engine Opticon System using a QuantiTect SYBR Green reverse transcription-PCR kit (Qiagen) under the conditions recommended by the manufacturer. Reverse transcription was done (50°C for 30 minutes) using a mixture (20 µL) consisting of 1 µL sample, 10 µL Master mix (Qiagen), 0.2 µL QuantiTect reverse transcription mix, and 1.5 µL (12.5 µmol/L) of each primer specific for the VEGF-D gene sequence (5'-ACAAGATGAGAATCCACTGCCTGG-3', 5'-CTCCAGGACATGGTGCTTTACAGA-3'). This was followed by initial denaturation, enzyme activation (95°C for 15 minutes), and 36 PCR cycles (94°C for 15 seconds, 53°C for 30 seconds, and 72°C for 30 seconds). The levels of mRNA for VEGF-D were normalized against the mRNA levels of ß-actin (5'-AGAAGATCTGGCACCACACC-3', 5'-CTCATCGTACTCCTGGTTGC-3') by calculating the difference in cycle threshold (C_T). The mRNA levels for VEGF-D in the VEGF^– and VEGF⁺ tumors were then compared with the mRNA levels for VEGF-D in the VEGF⁰ tumors by calculating the difference in C_T (C_T). The mRNA levels were then expressed as fold increase (2^–C_T) as per standard comparative reverse transcription-PCR methods (PE Applied Biosystems User Bulletin 2).

Magnetic resonance imaging. A 7 T MRI system (Varian, Inc., Palo Alto, CA) was used to acquire all images. During all MRI procedures, animals were anesthetized with an isoflurane (1.5% in air) and the body temperature was maintained at 37°C using a heated water recirculating pad.

A fast spin echo multislice imaging sequence was used (4-second TR, 60 ms TE_eff, 128 x 128 matrix, 13 x 1-mm-thick slices, 8 echo train length, and a 3 cm field of view) to image the animals every 2 days. From these images, tumor volumes were determined by drawing tumor regions of interest in each slice using a drawing tool developed with Matlab. Volumes from all slices for an individual animal were summed at each time point and the natural logarithms of these volumes were plotted as a function of time postimplantation to calculate tumor growth rates. In addition, as a measure of tumor latency, the times for tumors to reach a volume of 20 µL (T₂₀) were interpolated from these plots.

A single-coil continuous arterial spin labeling imaging protocol (14) was used to obtain blood flow maps of the orthotopic tumors in vivo. Briefly, continuous arterial spin labeling perfusion and nonperfusion weighted images were acquired using a modified fast spin echo imaging sequence (4-second TR, 15 ms TE_eff, 128 x 128 matrix, 1.5 mm slice thickness, 8 echo train length, and a 3 cm field of view). The perfusion weighting was accomplished with a train of hyperbolic secant inversion pulses to invert the inflowing blood spins preceding the image acquisition. Blood flow images were then calculated on a pixel-by-pixel basis from these perfusion images using the following equation:

(1)

where F is the blood flow (mL-s/100 g-min); s and s₀ are the perfusion and nonperfusion weighted signal intensities, respectively; is the labeling efficiency; is the blood partition coefficient; and T₁ is the effective spin-lattice relaxation time of the water contained within the tissue. Maps of T₁ were calculated from a series of five inversion recovery fast spin echo T₁ weighted images (50 ms to 5-second inversion time, 1-second TR, 15 ms TE_eff, 1.5 mm slice thickness, 16 echo train length, 128 x 64 matrix, and a 3 cm field of view). A nonlinear least-squares fitting algorithm (Levenberg-Marquardt) was used to fit for T₁ on a pixel-by-pixel basis.

Relative cerebral blood volume (rCBV) maps were acquired using a gradient echo T₂* weighted MRI protocol. Feridex IV (15 mg Fe/kg; Berlex, Montville, NJ) was administered via the tail vein over a 30-second period. Using a gradient echo MRI sequence, T₂* weighted images (10 ms TR, 5 ms TE, 1 mm slice thickness, 128 x 128 matrix, and a 3 cm field of view) were acquired before and after injection (within 5 minutes) through the middle of the tumor. The change in T₂* relaxivity (R₂*) is proportional to blood volume. Therefore, the rCBV maps were calculated on a pixel-by-pixel basis using the following formula:

(2)

where R₂*_av is the mean change in relaxivity of the contralateral brain.

Diffusion MRI was done to obtain maps of tumor apparent diffusion coefficient at baseline and posttherapy using the previously described method (15). Briefly, a trace diffusion weighted multislice spin echo sequence (with motion compensation and navigator echo) was used to acquire images (3-second TR, 60 ms TE_eff, 13 x 1-mm-thick slices, 128 x 128 image matrix, and a 30 mm field of view) with two different diffusion weightings (b value of 1,148 s/mm²).

Treatment with BCNU. In an additional set of animals (n = 5 per tumor line), response to BCNU treatment was investigated. When in vivo tumor volumes were between 40 and 60 µL, BCNU (13.3 mg/kg) was administered in 10% ethanol by a single i.p. injection. Diffusion and T₂ weighted MRI was done every other day posttherapy to measure changes in tumor cell density and tumor volume, respectively. Animals were euthanized when discharge appeared around both eyes. Animal survival time was defined as the time from tumor treatment to 1 day after euthanasia.

Statistical analysis. Linear least-squares analysis was used to calculate the tumor growth rates from the semilog plots of tumor volume as a function of time postimplantation. The T₂₀ values were interpolated from the calculated line of best fit. When results from two groups were compared, a two-tailed Student's t test was used. Both linear least squares and Student's t test were done using Microsoft Excel. A log-rank test was done (Prism, GraphPad Software, Inc., San Diego, CA) to test the statistical significance of differences in animal survival.

	Results

Top Abstract Materials and Methods Results Discussion References

 
In vitro characterization of the 9L cell lines. Two new stable (VEGF^– and VEGF⁺) 9L brain tumor cell lines were produced. The VEGF-A expression of both these cell lines (Fig. 1) revealed different profiles of VEGF-A expression. The VEGF^– cells had no detectable amounts of VEGF-A (Fig. 1, lane 2), whereas the VEGF⁺ cells had a significantly higher level of VEGF-A expression (Fig. 1, lane 1) compared with the VEGF⁰ cell line (Fig. 1, lane 3). In addition, the growth rates of the cell lines were measured to determine if changing VEGF-A expression affected the intrinsic growth rate of the 9L tumor cells in culture. As shown in Table 1, the VEGF^– cells had a slightly slower cell growth rate compared with VEGF⁰ cells, whereas the VEGF⁺ cells had a slightly faster growth rate. These changes in growth rate also correlated with slight differences in sensitivity to BCNU as determined using the sulfur rhodamine-B growth inhibition assay (see IC₅₀ values in Table 1).

View larger version (13K): [in this window] [in a new window]   Fig. 1. In vitro analysis of three 9L cell lines with different levels of VEGF-A expression. Western analysis of in vitro expression of VEGF-A in the VEGF+ (lane 1), VEGF– (lane 2), and VEGF0 (lane 3) cell lines.

View larger version (43K): [in this window] [in a new window]   Fig. 2. In vivo growth characteristics of 9L orthotopic tumors with different levels of VEGF-A expression. A, anatomic T2 weighted MRIs of rat brains harboring VEGF–, VEGF0, and VEGF+ tumors 20 days postimplantation. The intracranial tumor volumes were 8, 140, and 322 µL, respectively. B, in vivo tumor doubling times (days, ± SE), which were calculated after tumors had achieved a minimum volume of 20 µL. C, survival curves for the animals implanted orthotopically with VEGF– (35 days), VEGF0 (26 days), and VEGF+ (20 days) cells (n = 6 per group). *, statistically different (with 95% confidence) from VEGF0 group.

Histopathology and immunohistochemistry. Pathologic sections from each tumor type were stained with H&E and examined under low and high magnification (Fig. 3A). All tumors were hypercellular, composed of predominantly spindled cells on a microcystic background. The amount of the cytoplasm was minimal to none in the cells of the VEGF⁰ and VEGF^– tumors, but quite visible in the VEGF⁺ tumors. Overall, cellularity was highest in the VEGF^– tumors due to frequent small rounded nuclei, most of which appeared pyknotic. The nuclei of the VEGF⁺ tumors were larger, with complex chromatin. Varying numbers of multinucleated cells were present in all tumors. The number of mitotic figures ranged from 4 to 11 per high-power field in the VEGF⁰ and VEGF⁺ tumors, to 3 to 9 in the VEGF^– tumors.

View larger version (79K): [in this window] [in a new window]   Fig. 3. Histopathology and MRI perfusion images of orthotopic 9L tumors with different levels of VEGF-A expression. A, representative H&E, anti-von Willebrand factor (vWF, factor VIII), and anti-VEGF-D–stained slices of VEGF–, VEGF0, and VEGF+ tumors, respectively. Arrows, positive VEGF-D staining. Magnification, x400. B, in vivo MRI tumor perfusion images showing representative rCBV and blood flow maps of VEGF–, VEGF0, and VEGF+ tumors.

Tumor blood volume and blood flow. To determine the effect of VEGF expression on tumor blood flow and blood volume, two independent MRI techniques were used to evaluate these variables. Cerebral blood volume and tumor blood flow maps were acquired once tumors were established (visible on at least three sequential T₂ weighted image slices). This enabled quantification and visualization of blood volume and flow heterogeneity within the tumors. Examples of tumor rCBV and blood flow maps (Fig. 3B) are shown. The VEGF^– tumors (Fig. 4A) had a mean rCBV (2 ± 1) that was not significantly (P = 0.85) different than that of the VEGF⁰ tumors (2 ± 1), whereas the VEGF⁺ tumors had a higher mean rCBV (8 ± 3, P < 0.001) compared with both the VEGF^– and VEGF⁰ tumors. The mean blood flow of these tumors (Fig. 4B) revealed VEGF⁺ tumors to have the greatest heterogeneity and highest mean blood flow (325 ± 24 mL/100 g/min), which was significantly (P < 0.001) greater than the VEGF⁰ tumors (43 ± 4 mL/100 g/min). The VEGF^– tumors had a mean blood flow (99 ± 9 mL/100 g/min) that was significantly (P < 0.001) higher than that of the VEGF⁰ tumors.

View larger version (12K): [in this window] [in a new window]   Fig. 4. Effects of VEGF-A expression levels on orthotopic 9L tumor perfusion. A, graph of the mean blood flow (±SE) of VEGF–, VEGF0, and VEGF+ tumors (n = 6 per group). B, graph of the mean rCBV (±SE) of VEGF–, VEGF0, and VEGF+ tumors. *, statistical difference (with 95% confidence) from VEGF0 group.

View larger version (49K): [in this window] [in a new window]   Fig. 5. Western analysis of VEGF-A and VEGF-D from excised tumor tissue. VEGF-A, full-length (53 kDa) VEGF-D, and mature form (23 kDa) VEGF-D expression of excised orthotopic VEGF+ (lane 1), VEGF– (lane 2), and VEGF0 (lane 3) tumors.

Treatment with BCNU. To investigate the effect of VEGF-A expression on tumor response to chemotherapy, the three 9L clones were tested for their sensitivity to BCNU both in vitro and in vivo. In vitro sensitivity was tested by comparing the IC₅₀ of BCNU (Table 1) for each cell line. The IC₅₀ values of both the VEGF^– and VEGF⁺ cell lines were not significantly different from that of the VEGF⁰ cell line.

For in vivo evaluation, animals were implanted with each 9L cell line and tumors were allowed to develop. To minimize the effect of tumor latency (see Fig. 2 and discussion above), tumors were allowed to reach an equivalent size (40-60 µL) before BCNU was administered to the animals (13.3 mg/kg). Following treatment with BCNU, the apparent diffusion coefficient increased significantly above baseline within 5 days of therapy (Fig. 6A). The average apparent diffusion coefficient increase in the VEGF⁰ tumors was larger than that observed for the VEGF⁺ and VEGF^– tumors; however, this difference did not reach statistical significance (P = 0.071 and 0.075, respectively). The median survival (40 days posttherapy) of animals with VEGF^– tumors was significantly (P = 0.02) longer than that for the animals with VEGF⁰ (33 days) tumors. Interestingly, this was not significantly (P = 0.7) longer than the survival of animals with VEGF⁺ tumors (36 days; Fig. 6B). Similarly, the difference in the median survival time between the VEGF⁰ and VEGF⁺ groups was not significant (P = 0.35; Fig. 6B). Tumor volumes obtained from serial MRI scans posttherapy (Fig. 6C) revealed that both the VEGF^– and the VEGF⁺ tumors had different volumetric response and regrowth rates compared with the VEGF⁰ tumors. The VEGF^– tumors had the slowest but most prolonged response to BCNU treatment reaching a nadir 22 days after BCNU treatment that was statistically (P = 0.02) longer than that of the VEGF⁰ tumors (16.2 days). The VEGF⁺ tumors responded the fastest such that they reached a nadir volume 13.9 days after treatment followed by rapid regrowth. This was found to be statistically different from the VEGF^– tumors (P = 0.02) but not in the VEGF⁰ tumors (P = 0.25). The nadir volumes for the VEGF^– (20%) and VEGF⁺ (28%) tumors were not statistically different (P = 0.73 and 0.52, respectively) from the VEGF⁰ (23%) tumors.

View larger version (25K): [in this window] [in a new window]   Fig. 6. Chemotherapy treatment results. A, mean apparent diffusion coefficient changes (±SE) 5 days post-BCNU (13.3 mg/kg) administration. Increased apparent diffusion coefficient indicates decreased viable tumor cell density. B, animal survival after time of treatment. VEGF– tumors had a significantly (P = 0.02, log-rank test) longer median survival time (40 days) than the VEGF0 tumors (33 days). However, the median survival of VEGF+ (36 days) and VEGF0 tumors were not statistically (P = 0.3, log-rank test) different. C, normalized intracranial tumor volumes (±SE) following treatment as measured by multislice T2 weighted MRI.

	Discussion

Top Abstract Materials and Methods Results Discussion References

Recent experimental results revealed that when antiangiogenic therapy is used in conjunction with cytotoxic therapies, a synergistic improvement in efficacy is achieved (25). This led Jain et al. (7) to postulate that the synergy may be due to the "paradox of antiangiogenic therapy" referring to observations that antiangiogenic therapies reduce blood vessel growth and permeability while also "normalizing" existing blood vessels resulting in a transient increase in blood flow and decrease tumor hypoxia. The increased blood flow allows for improved drug penetration whereas the reduction in tumor hypoxia improves the efficacy of cytotoxic therapies, such as ionizing radiation (25). To maximize the synergistic effects of combined antiangiogenic and cytotoxic therapies, the development of noninvasive biomarkers of blood flow and tumor oxygenation is extremely important. In this study, we have shown that blood flow as measured by MRI has the potential to be an important biomarker of the normalization of blood vessels during antiangiogenic therapies. The results show that modulation of VEGF-A production resulted in significantly different blood flow properties of orthotopic 9L tumors. Once established, the VEGF^– tumors had a higher mean blood flow and blood vessel densities than did the VEGF⁰ tumors. Interestingly, the relative blood volume of these tumors was not significantly different from the VEGF⁰ tumors. These observations suggest that VEGF-A suppression leads to a more efficient blood vasculature, consistent with the "normalization" theory of Jain et al. (7). It is important to note that the blood flow measurements made using the continuous arterial spin labeling method (14) were completely independent measurements of perfusion from the blood volume maps and different to other contrast enhanced measurements of tumor perfusion (26–32).

This study also revealed that restricting the expression of VEGF-A by the tumor cells is not sufficient to block angiogenesis in 9L tumors. The observation that the VEGF^– tumors could overcome the VEGF-A deficiency by forming new blood vessels, thereby attaining greater blood flow than parental 9L tumors, suggested the involvement of an alternative pathway or different angiogenic molecule. There are a number of possible explanations for this observation. One possibility is that the VEGF^– tumors escaped VEGF-A suppression similar to a previous report (21). However, Fig. 5 shows that the VEGF^– tumors have suppressed VEGF-A expression relative to the VEGF⁰ tumors. Another study (33) suggested that stromal production of VEGF-A may contribute significantly to the growth of tumors in rodent tumor models. In that study, maximum antiangiogenic effect occurred when both tumor and host production of VEGF-A was suppressed. This latter fact may explain the higher expression of VEGF-A in VEGF^– tumors compared with the expression of VEGF-A in VEGF^– cell cultures that was observed in this study. The Western blots and quantitative PCR, however, both identified up-regulation of VEGF-D in the VEGF^– tumors compared with that of both the VEGF⁰ and VEGF⁺ tumors. These results suggest that VEGF-D is a likely mediator of angiogenesis in these tumors with negligible VEGF-A expression. VEGF-D is a well-known angiogenic (16, 17) and lymphangiogenic (34) growth factor that is most likely regulated by cell-to-cell contact (35). VEGF-A and VEGF-D are also both highly expressed in glioblastoma (2). Considering the angiogenic ability of VEGF-D, it is perhaps not surprising to observe that in tumors in which the availability of VEGF-A was restricted, the tumors were subsequently able to up-regulate VEGF-D. In accordance with this theory, current pharmacologic research and recent clinical trials in patients with high-grade glioma have revealed enhanced suppression of angiogenesis and a clinical benefit through the inhibition of all receptors within the VEGF family (36, 37).

When treated with BCNU, the different 9L tumor types showed different response rates. However, this did not translate into significant differences in apparent diffusion coefficient change and nadir volume although the time to nadir and animal survival were significantly longer for the VEGF^– tumor compared with the VEGF⁰ tumors. The increase in survival following BCNU treatment of the VEGF^– tumors can largely be accounted for by the slower doubling time of the VEGF^– tumors and does not support an increase in BCNU sensitivity as the major factor in this prolonged survival. The effect of tumor growth rate on survival following BCNU treatment can also be seen when comparing the survival of animals bearing the different VEGF-A expressing tumors that were treated (Fig. 6B) or not treated (Fig. 2C). The differences in survival between the groups were similar whether they received treatment or not; however, the overall median survival was increased for each group with BCNU treatment.

In conclusion, altering the expression levels of VEGF-A in the 9L tumor model resulted in significant changes in tumor perfusion. Both the 9L tumor models overexpressing and underexpressing VEGF-A exhibited elevated tumor blood flow. Blood volume was elevated in the overexpressing VEGF-A tumors, but was not significantly altered in underexpressing tumors. In addition, when VEGF-A was inhibited, there was an increased latency period until tumors developed; however, once they were established, they developed a more efficient blood supply than the wild-type tumors. Subsequent analysis showed that the development of these blood vessels may have occurred as a result of VEGF-D up-regulation. These results show the "adaptive ability" of tumor cells to antiangiogenic therapy. Not only do tumors have multiple pathways for angiogenic signaling, but they also have the ability to call on these secondary pathways when the primary pathway is inhibited. It should also be noted that for brain tumors the up regulation of VEGF-D and its potential for lymphangiogenesis may not be a concern due to the negligible amount of lymph vessels in the brain for co-option. However, for antiangiogenic treatments targeting only VEGF-A in tumors located where lymphatic vessels can be co-opted, these results raise the possibility of increased lymphatic spread and metastatic potential.

These results show that antiangiogenic treatments may be used effectively in combination with chemotherapy and/or radiotherapy (38). Therefore, finding the optimal combination, dose, and timing of these therapies for maximizing the synergistic effects poses important challenges for cancer therapy. To confront these challenges, it is becoming clear that imaging biomarkers of antiangiogenic therapeutic efficacy are extremely important. The results of this study show that the clinically translatable MRI blood flow and blood volume maps are independent imaging strategies that can provide complementary and important information with respect to tumor angiogenesis.

	Footnotes

 
Grant support: NIH/National Cancer Institute grants PO1CA85878, R24CA83099, and P50CA093990; and the support of the tissue, vector, and microarray cores within the Michigan Comprehensive Cancer Center (5P30CA46592).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 6/28/05; revised 11/29/05; accepted 12/16/05.

	References

Top Abstract Materials and Methods Results Discussion References

 

1. Folkman J. Tumor angiogenesis: therapeutic implications. N Engl J Med 1971;285:1182–6.[Medline]
2. Jansen M, de Witt Hamer PC, Witmer AN, Troost D, van Noorden CJ. Current perspectives on antiangiogenesis strategies in the treatment of malignant gliomas. Brain Res Brain Res Rev 2004;45:143–63.[CrossRef][Medline]
3. Rehman S, Jayson GC. Molecular imaging of antiangiogenic agents. Oncologist 2005;10:92–103.[Abstract/Free Full Text]
4. Rajendran JG, Krohn KA. Imaging hypoxia and angiogenesis in tumors. Radiol Clin North Am 2005;43:169–87.[CrossRef][Medline]
5. Miller JC, Pien HH, Sahani D, Sorensen AG, Thrall JH. Imaging angiogenesis: applications and potential for drug development. J Natl Cancer Inst 2005;97:172–87.[Abstract/Free Full Text]
6. Jain RK. Molecular regulation of vessel maturation. Nat Med 2003;9:685–93.[CrossRef][Medline]
7. Jain RK. Normalization of tumor vasculature: an emerging concept in antiangiogenic therapy. Science 2005;307:58–62.[Abstract/Free Full Text]
8. Risau W. Mechanisms of angiogenesis. Nature 1997;386:671–4.[CrossRef][Medline]
9. Plate KH, Breier G, Millauer B, Ullrich A, Risau W. Up-regulation of vascular endothelial growth factor and its cognate receptors in a rat glioma model of tumor angiogenesis. Cancer Res 1993;53:5822–7.[Abstract/Free Full Text]
10. Nam DH, Park K, Suh YL, Kim JH. Expression of VEGF and brain specific angiogenesis inhibitor-1 in glioblastoma: prognostic significance. Oncol Rep 2004;11:863–9.[Medline]
11. Peles E, Lidar Z, Simon AJ, et al. Angiogenic factors in the cerebrospinal fluid of patients with astrocytic brain tumors. Neurosurgery 2004;55:562–7; discussion 567–8.[CrossRef][Medline]
12. Conn G, Bayne ML, Soderman DD, et al. Amino acid and cDNA sequences of a vascular endothelial cell mitogen that is homologous to platelet-derived growth factor. Proc Natl Acad Sci U S A 1990;87:2628–32.[Abstract/Free Full Text]
13. Skehan P, Storeng R, Scudiero D, et al. New colorimetric cytotoxicity assay for anticancer-drug screening. J Natl Cancer Inst 1990;82:1107–12.[Abstract/Free Full Text]
14. Moffat BA, Chenevert TL, Hall DE, Rehemtulla A, Ross BD. Continuous arterial spin labeling using a train of adiabatic inversion pulses. J Maga Reson Imaging 2005;21:290–6.[CrossRef]
15. Moffat BA, Hall DE, Stojanovska J, et al. Diffusion imaging for evaluation of tumor therapies in preclinical animal models. MAGMA 2004;17:249–59.[CrossRef][Medline]
16. Achen MG, Jeltsch M, Kukk E, et al. Vascular endothelial growth factor D (VEGF-D) is a ligand for the tyrosine kinases VEGF receptor 2 (Flk1) and VEGF receptor 3 (Flt4). Proc Natl Acad Sci U S A 1998;95:548–53.[Abstract/Free Full Text]
17. Marconcini L, Marchio S, Morbidelli L, et al. c-fos-induced growth factor/vascular endothelial growth factor D induces angiogenesis in vivo and in vitro. Proc Natl Acad Sci U S A 1999;96:9671–6.[Abstract/Free Full Text]
18. Stacker SA, Stenvers K, Caesar C, et al. Biosynthesis of vascular endothelial growth factor-D involves proteolytic processing which generates non-covalent homodimers. J Biol Chem 1999;274:32127–36.[Abstract/Free Full Text]
19. Debinski W, Slagle-Webb B, Achen MG, et al. VEGF-D is an X-linked/AP-1 regulated putative onco-angiogen in human glioblastoma multiforme. Mol Med 2001;7:598–608.[Medline]
20. Sasaki M, Wizigmann-Voos S, Risau W, Plate KH. Retrovirus producer cells encoding antisense VEGF prolong survival of rats with intracranial GS9L gliomas. Int J Dev Neurosci 1999;17:579–91.[CrossRef][Medline]
21. Cheng SY, Huang HJ, Nagane M, et al. Suppression of glioblastoma angiogenicity and tumorigenicity by inhibition of endogenous expression of vascular endothelial growth factor. Proc Natl Acad Sci U S A 1996;93:8502–7.[Abstract/Free Full Text]
22. Hagedorn M, Javerzat S, Gilges D, et al. Accessing key steps of human tumor progression in vivo by using an avian embryo model. Proc Natl Acad Sci U S A 2005;102:1643–8.[Abstract/Free Full Text]
23. Alon T, Hemo I, Itin A, et al. Vascular endothelial growth factor acts as a survival factor for newly formed retinal vessels and has implications for retinopathy of prematurity. Nat Med 1995;1:1024–8.[CrossRef][Medline]
24. Oku T, Tjuvajev JG, Miyagawa T, et al. Tumor growth modulation by sense and antisense vascular endothelial growth factor gene expression: effects on angiogenesis, vascular permeability, blood volume, blood flow, fluorodeoxyglucose uptake, and proliferation of human melanoma intracerebral xenografts. Cancer Res 1998;58:4185–92.[Abstract/Free Full Text]
25. Winkler F, Kozin SV, Tong RT, et al. Kinetics of vascular normalization by VEGFR2 blockade governs brain tumor response to radiation: role of oxygenation, angiopoietin-1, and matrix metalloproteinases. Cancer Cell 2004;6:553–63.[Medline]
26. Kobayashi H, Sato N, Hiraga A, et al. 3D-micro-MR angiography of mice using macromolecular MR contrast agents with polyamidoamine dendrimer core with reference to their pharmacokinetic properties. Magn Reson Med 2001;45:454–60.[CrossRef][Medline]
27. Neeman M, Provenzale JM, Dewhirst MW. Magnetic resonance imaging applications in the evaluation of tumor angiogenesis. Semin Radiat Oncol 2001;11:70–82.[CrossRef][Medline]
28. Abramovitch R, Dafni H, Smouha E, Benjamin LE, Neeman M. In vivo prediction of vascular susceptibility to vascular susceptibility endothelial growth factor withdrawal: magnetic resonance imaging of C6 rat glioma in nude mice. Cancer Res 1999;59:5012–6.[Abstract/Free Full Text]
29. Akella NS, Twieg DB, Mikkelsen T, et al. Assessment of brain tumor angiogenesis inhibitors using perfusion magnetic resonance imaging: quality and analysis results of a phase I trial. J Magn Reson Imaging 2004;20:913–22.[CrossRef][Medline]
30. Brasch R, Turetschek K. MRI characterization of tumors and grading angiogenesis using macromolecular contrast media: status report. Eur J Radiol 2000;34:148–55.[CrossRef][Medline]
31. George ML, Dzik-Jurasz AS, Padhani AR, et al. Non-invasive methods of assessing angiogenesis and their value in predicting response to treatment in colorectal cancer. Br J Surg 2001;88:1628–36.[CrossRef][Medline]
32. Tropres I, Grimault S, Vaeth A, et al. Vessel size imaging. Magn Reson Med 2001;45:397–408.[CrossRef][Medline]
33. Gerber HP, Kowalski J, Sherman D, Eberhard DA, Ferrara N. Complete inhibition of rhabdomyosarcoma xenograft growth and neovascularization requires blockade of both tumor and host vascular endothelial growth factor. Cancer Res 2000;60:6253–8.[Abstract/Free Full Text]
34. Stacker SA, Caesar C, Baldwin ME, et al. VEGF-D promotes the metastatic spread of tumor cells via the lymphatics. Nat Med 2001;7:186–91.[CrossRef][Medline]
35. Orlandini M, Marconcini L, Ferruzzi R, Oliviero S. Identification of a c-fos-induced gene that is related to the platelet-derived growth factor/vascular endothelial growth factor family. Proc Natl Acad Sci U S A 1996;93:11675–80.[Abstract/Free Full Text]
36. Goldbrunner RH, Bendszus M, Wood J, et al. PTK787/ZK222584, an inhibitor of vascular endothelial growth factor receptor tyrosine kinases, decreases glioma growth and vascularization. Neurosurgery 2004;55:426–32; discussion 432.[CrossRef][Medline]
37. Conrad C, Friedman H, Reardon D, et al. A Phase I/II trial of single-agent PTK 787/ZK 222584 (PTK/ZK), a novel, oral angiogenesis inhibitor, in patients with recurrent glioblastoma multiforme (GBM). J Clin Oncol 2004;22:1512.
38. Willett CG, Boucher Y, di Tomaso E, et al. Direct evidence that the VEGF-specific antibody bevacizumab has antivascular effects in human rectal cancer. Nat Med 2004;10:145–7.[CrossRef][Medline]

This article has been cited by other articles:

				S. Senan and E. F. Smit Design of Clinical Trials of Radiation Combined with Antiangiogenic Therapy Oncologist, April 1, 2007; 12(4): 465 - 477. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Moffat, B. A. Articles by Ross, B. D. Search for Related Content PubMed PubMed Citation Articles by Moffat, B. A. Articles by Ross, B. D.