Vascular Endothelial Growth Factor mRNA Expression and Peritumoral Edema in Canine Primary Central Nervous System Tumors

Abstract

Vascular endothelial growth factor (VEGF) is an important regulator of tumor angiogenesis and vascular permeability, and has been implicated both in progression of central nervous system (CNS) tumors and development of vasogenic peritumoral edema. A retrospective study was done to characterize the levels of expression of the 3 major canine VEGF isoforms (VEGF₁₂₀, VEGF₁₆₄, VEGF₁₈₈) in a variety of spontaneous canine CNS tumors using quantitative TaqMan reverse transcription real-time polymerase chain reaction. Presence and degree of peritumoral edema also were determined in sampled tumors using magnetic resonance imaging (MRI). Increased expression of VEGF relative to normal cerebral cortex tissue was seen predominantly in high grade astrocytic (grade IV) and oligodendroglial (grade III) tumors, with lower expression in low grade astrocytomas (grade II) and meningiomas (grade I). All 3 major VEGF isoforms were present; VEGF₁₆₄ was the predominant isoform, particularly in the tumors with the highest VEGF expression. Peritumoral edema was present in all tumor types; however, a significant association between the extent of peritumoral edema and the level of VEGF expression was not apparent.

Keywords

Canine central nervous system tumors edema reverse transcriptase polymerase chain reaction vascular endothelial growth factor VEGF

Introduction

Vascular endothelial growth factor (VEGF), also termed VEGFA and vascular permeability factor, is a member of a family of angiogenesis-related growth factors. It is a potent endothelial cell mitogen and survival factor, and a strong inducer of vascular permeability, a thousand times more potent than histamine.^17, ²² It is an essential factor for angiogenesis and vasculogenesis during normal development and is proposed to play a major role in tumor neovascularization and associated edema.^11, ²⁸ The VEGF gene consists of 8 exons, and in humans at least 9 VEGF isoforms are produced by alternative splicing: VEGF₁₂₁, VEGF₁₄₅, VEGF₁₄₈, VEGF₁₆₂, VEGF₁₆₅, VEGF_165b, VEGF₁₈₃, VEGF₁₈₉, VEGF₂₀₆.^2, ^10, ^11, ⁴⁸ VEGF₁₂₁, VEGF₁₆₅, and VEGF₁₈₉ are the major isoforms secreted by most cell types, with VEGF₁₆₅ being the most abundant isoform in both normal and transformed cells.^10, ⁴⁴ The major VEGF isoforms bind cell-surface heparan sulphate proteoglycans to varying degrees, with VEGF₁₂₁ being freely diffusible, approximately half of secreted VEGF₁₆₅ being bound, and almost all VEGF₁₈₉ remaining sequestered in the extracellular matrix.

VEGF is overexpressed at both the mRNA and protein level in a variety of human brain tumors, including astrocytomas,^13, ^14, ^29, ^33, ^38, ^39, ^45, ^49, ⁵⁴ oligodendrogliomas,^6, ^7, ^14, ^19, ^45, ⁵² and meningiomas.^13, ^14, ^21, ^36, ^45, ⁴⁹ VEGF expression has been associated with a variety of factors, including necrosis and hypoxia, growth factors such as transforming growth factor (TGF) and fibroblast growth factor (FGF), cytokines including IL-1α and IL-6, and oncogenes or tumor suppressor genes such as Ras and TP53.^11, ²⁸ Most studies have linked the degree of overexpression of VEGF with increasing tumor grade in astrocytomas, oligodendrogliomas, and meningiomas;^5, ^6, ^13, ^14, ^19, ^21, ^29, ^36, ^38, ^50, ^52, ⁵⁴ however, this has not been universally reported.^35, ^45, ⁴⁷

Vasogenic peritumoral edema is a well-documented complication contributing to morbidity and death in patients with central nervous system (CNS) tumors. The exact mechanisms responsible for increased vascular permeability in specific tumors have not been elucidated fully and are likely to be multifactorial; however, VEGF has been implicated as a contributory agent in tumor-related edema in humans. Peritumoral edema, increased vascular permeability, and presence of intratumoral cysts have been correlated with increased VEGF expression in human gliomas^20, ^26, ^47, ⁵¹ and meningiomas.^12, ^16, ^34, ^35, ^37, ⁴¹ In addition, alleviation of peritumoral edema in high grade gliomas following administration of VEGF-receptor-specific tyrosine kinase inhibitors indirectly suggests an important role for VEGF in edema formation.¹

Five VEGF isoforms have been reported in dogs: VEGF₁₂₀, VEGF₁₄₄, VEGF₁₆₄, VEGF₁₈₂, and VEGF₁₈₈. These isoforms have approximately 95% homology to the human amino acid sequence and, similar to most other mammals, lack a single amino acid residue in the N-terminal region compared with human VEGF.^15, ⁴⁶ VEGF expression has been demonstrated by means of immunohistochemistry and polymerase chain reaction (PCR) and in a variety of canine neoplasms,^30, ^42, ^43, ⁴⁶ including 2 canine grade IV astrocytomas²³ and a series of grade 1 meningiomas.⁴⁰ Few data exist describing the relative expression of VEGF isoforms in canine tissues, although all available data suggest that VEGF₁₂₀, VEGF₁₆₄, and VEGF₁₈₈ are the most abundant isoforms, as in other mammalian species; VEGF₁₂₀, VEGF₁₆₄, and VEGF₁₈₈ were shown by RT-PCR and sequencing to be the major isoforms in canine lung carcinoma,⁴⁶ and similar data were reported in canine heart using RT-PCR, ribonuclease protection assays, and Western blotting.¹⁵ The objectives of this study were twofold: 1) to evaluate the mRNA expression of the major VEGF isoforms in a variety of primary canine CNS tumors using quantitative real-time TaqMan polymerase chain reaction and to correlate any mRNA expression with tumor type and grade, and 2) to determine whether there was an association between the level of VEGF mRNA expression in canine CNS tumors and the presence or severity of peritumoral edema evaluated by magnetic resonance imaging (MRI).

Materials and Methods

Sample collection

All tumor tissue was obtained from surgical biopsy/resection specimens, necropsy, or from archival paraffin-embedded material from clinical cases presented to the Veterinary Medical Teaching Hospital, University of California, Davis. Samples of tumor from necropsy were collected within 20 minutes of death and snap frozen in liquid nitrogen. Surgical samples were similarly stored following collection. Samples of adjacent tumor tissue were processed for routine paraffin embedding and histology whenever fresh tissue was collected in liquid nitrogen. Normal brain samples were collected from both necropsy and archival paraffin-embedded material. All tumors were graded by a board certified pathologist (RJH, AWB) essentially according to the international World Health Organization classification of human tumors of the nervous system.^18, ^24, ²⁵ Meningiomas were graded as either grade I (benign), grade II (atypical), or grade III (malignant); astrocytomas were graded as either grade II (diffuse), grade III (anaplastic), or grade IV (glioblastoma multiforme); oligodendroglial tumors were graded as either grade II (oligodendroglioma) or grade III (anaplastic oligodendroglioma). Clinical records were reviewed for evidence of corticosteroid administration prior to either MRI or acquisition of tumor samples.

Magnetic resonance imaging

MRI was done using either a 0.4T scanner (Resonex 5000, Resonex Inc., Sunnyvale, CA) or a 1.5T scanner (General Electric, Sigma, Milwaukee, WI). The imaging sequences were obtained from transverse 5-mm slices with an interslice gap of 1 mm generated with T1-weighted (T1W), T2-weighted (T2W), and proton-density weighted (PDW) spin echo pulse sequences. T1-weighted images were also acquired after the intravenous administration of gadolinium-diethylenetriaminepenta-acetic acid (Gd-DTPA; Magnevist, Berlex Laboratories, Wayne, NJ) at 0.2 mmol/kg. Peritumoral edema was determined to be present or absent based on the presence of hyperintensity on T2-weighted images beyond the tumor boundary. Tumor boundaries were determined based on T1W, T1W postcontrast, T2W, and PDW images. Edema was graded either as absent, focal if edema was present on 1 or 2 consecutive T2W images beyond the tumor margin, or diffuse if present on 3 or more consecutive images beyond the tumor margin.

Canine VEGF sequence

Canine VEGF cDNA sequence data and clones for VEGF isoforms VEGF₁₂₀, VEGF₁₆₄, and VEGF₁₈₈ were kindly provided by Dr. Rolf Jaussi (Institute of Medical Radiobiology of the University of Zurich and the Paul Scherrer Institute, Zurich, Switzerland).

TaqMan PCR

Total RNA extraction, cDNA preparation, and real-time TaqMan PCR were done as previously described.⁹ VEGF PCR primers for canine isoforms VEGF₁₂₀, VEGF₁₄₄, and VEGF₁₈₈ were designed based on canine sequence data described above using Primer Express software (Applied Biosystems, Foster City, CA) and are listed and illustrated in Table 1 and Fig. 1. TaqMan PCR primers for canine housekeeping genes glyceraldehydes-3-phosphate dehydrogenase (GAPDH), ribosomal protein L13A, glycosyltransferase (HPRT1), and glucuronidase beta (GUSB) (Table 1) were designed and validated as previously described.⁹ PCR products were designed to be less than 150 base pairs in length, with either one of the primer pairs or internal probe placed over an exon-exon junction to allow discrimination between cDNA and gDNA. PCR products were initially validated by agarose gel electrophoresis. Subsequently, products from triplicate PCR reactions, generated using control canine pooled tissue cDNA and each primer pair, were combined and purified (QIAquick PCR purification kit, QIAGEN Inc., Valencia, CA). Specificity of the PCR product was then confirmed by direct sequencing using an automated big dye termination sequencing procedure (Davis Sequencing, Davis, CA). Transcript quantitation was done using the comparative C_T method and reported as relative transcription, or the n-fold difference relative to the mean value for individual normal cerebral cortex samples (n = 15). Tumor samples that had GAPDH C_T values weaker than 3 times the average GAPDH C_T value were considered low-quality cDNA samples and were discarded.

Table 1

TaqMan PCR primers.


Target	Forward Primer Sequence (5′→3′)	Reverse Primer Sequence (5′→3′)

GAPDH	GATGGGCGTGAACCATGAG	TCATGAGGCCCTCCACGAT
Ribosomal protein L13A	AAGTTGAATTACCTGGCCTTCCT	GGGTGCCCGGAAGTCATAG
Glycosyltransferase	GAGATGACCTCTCAACTTTAACTGAAAA	CAAGGGAAGCAAGGTTTGCA
B-glucuronidase	GCTGGATCAGAAACGCAAAGA	CTCTCTGTGGTGACTGGTCAGTCAT
VEGF₁₂₀	AGATTATGCGGATCAAACCTCATC	TCGGCTTGTCACATTTTTCTTG
VEGF₁₆₄	AGATTATGCGGATCAAACCTCATC	CCACAGGGATTTTCTTGCCTT
VEGF₁₈₈	AGATTATGCGGATCAAACCTCATC	CTCGAACTGATTTTTTTTCTTGCCTT
131		CCCTCAAGATTGTCAGCAATGCCTCCT
78		AACACCAACCCATCCCGTGGCC
92		CTTGATTGTTGAAGATCTCATTGACACAGGCA
86		TGGTTGGAGAGCTCATCTGGAATTTTGCT
123		CAGCATAGCAAATGTGAATGCAGACCAAA
118		CAGCATAGCAAATGTGAATGCAGACCAAA
123		CAGCATAGCAAATGTGAATGCAGACCAAA

Fig. 1.

Schematic representation of the exon structure and splice variants of canine and human VEGF. Known canine splice variants are in bold marked with an asterisk (∗). Isoform lengths are numbered as for canine. The canine VEGF primers and probes were designed to amplify canine VEGF isoforms 120, 164, and 188 using a common forward primer (F), a common internal probe (P), and reverse primers containing a common terminus (Rc) located at the end of exon 5 linked to isoform-specific sequence in exon 8 (VEGF₁₂₀), exon 7 (VEGF₁₆₄), and exon 6 (VEGF₁₈₈). It is unknown whether additional VEGF isoforms reported in humans are present in dogs. It is possible that the VEGF₁₆₄ primers might also amplify VEGF_164b and VEGF₁₄₇, and the VEGF₁₈₈ primers might amplify VEGF₁₈₂, VEGF₁₆₂, VEGF₁₄₅, and VEGF₂₀₅. Ex = exon, UTR = untranslated region. Sequence specificity was confirmed by direct sequencing.

Statistical analysis

Data were divided into nominal categories of tumor type, and for astrocytomas, they were further divided into ordinal categories of grade. Kruskal-Wallis one-way analysis of variance (ANOVA) was used to compare the variation in expression of VEGF mRNA among tumor types. When significant differences were evident, Mann-Whitney tests were used for pairwise comparisons. A sequentially rejective modification of Bonferroni's multiple comparison adjustment was used to confirm significant results. Variation in expression of specific VEGF isoforms within tumor groups was assessed using Friedman two-way ANOVA and Wilcoxon signed rank test for paired data. Peritumoral edema data were divided into ordinal categories of severity: 0 = no edema, 1 = focal edema, 2 = diffuse edema. Trends in expression of VEGF isoforms with severity of peritumoral edema were evaluated using the Jonckheere-Terpstra test. Evaluation of VEGF expression with presence or absence of edema was done using Mann-Whitney tests. The effect of presence or absence of corticosteroid administration on total VEGF expression and presence or absence of edema was assessed using Mann-Whitney tests. Statistical significance was defined as P ≤ .05.

Results

VEGF expression

PCR products from the 3 VEGF primer sets were of the appropriate size based on agarose gel electrophoresis, and direct sequencing of gel-extracted product resulted in single primer specific sequences corresponding to the 3 major isoforms, VEGF₁₂₀, VEGF₁₆₄, and VEGF₁₈₈. VEGF PCR product was amplified from 56 separate samples from 56 dogs; 15 samples of normal cerebral cortex, 9 grade I meningiomas, 1 grade II (atypical) meningioma, 6 grade II astrocytomas, 5 grade IV astrocytomas (glioblastoma multiforme [GBM]), 2 grade II oligodendrogliomas, and 18 grade III (anaplastic) oligodendrogliomas (Fig. 2). Three grade I meningiomas and 1 grade III oligodendroglioma were of spinal cord origin. Averages of normalized values and the standard deviations for both frozen and paraffin-embedded samples were not significantly different.

Fig. 2.

Expression of the major canine VEGF mRNA isoforms (a) and total VEGF mRNA isoforms (b) in primary canine CNS tumors. The highest expression is generally in high grade gliomas, and VEGF₁₆₄ is the predominant isoform present. MEN I = meningiomas; AST II = astrocytoma grade II; GBM = glioblastoma multiforme (astrocytoma grade IV); OLIGO III = oligodendroglioma grade III; 120 = VEGF₁₂₀ isoform mRNA; 164 = VEGF₁₆₄ isoform mRNA; 188 = VEGF₁₈₈ isoform mRNA. The top and bottom edge of the box plots represent the upper and lower quartile, respectively. The line within the box represents the median. The tails extend to the farthest point that is within 1.5 interquartile ranges of the quartiles. x-Axis divisions are proportional to sample size.

Increased expression of VEGF transcripts was seen most often with high grade gliomas, including GBMs and oligodendrogliomas. Combined expression of all isoforms was significantly greater in GBMs than grade II astrocytomas (P = .026) and significantly greater in oligodendrogliomas than meningiomas (P = .009) or grade II astrocytomas (P = .001). There was no significant difference between total VEGF expression in GBMs and oligodendrogliomas (P = .25) (Fig. 2b). Expression of VEGF isoforms in the single atypical meningioma was similar to that in the grade I meningiomas. Expression of VEGF isoforms in the 2 grade II oligodendrogliomas were ranked the second and fifth lowest of all oligodendroglial samples as a single group. All 3 major VEGF isoforms were present in tumor samples, and VEGF₁₆₄ was the predominant isoform in grade III oligodendrogliomas (P = .016) (Fig. 2a).

Information relating to administration of corticosteroids was available for 38/41 tumors. Corticosteroids of varying dose and duration were administered in 23/38 dogs prior to collection of tumor samples. There was no statistical difference in VEGF transcript levels in treated and untreated tumor samples when looking at individual tumor groups. Total VEGF expression with administration of corticosteroids approached significance when looking at all tumor groups combined (P = .055).

Peritumoral edema

MRI was available for 37 dogs from which tumor samples were analyzed. This included 9 meningiomas, 4 grade II astrocytomas, 5 GBMs, 2 grade II oligodendrogliomas, and 17 grade III oligodendrogliomas. Edema was absent in 2/9 meningiomas, 1/4 grade II astrocytoma, 2/2 grade II oligodendrogliomas, and 5/17 grade III oligodendrogliomas; focal in 4/9 meningiomas, 2/4 grade II astrocytomas, 5/5 GBMs, and 11/17 grade III oligodendrogliomas; diffuse in 3/9 meningiomas, 2/4 grade II astrocytomas, and 1/17 grade III oligodendrogliomas. No significant association was seen between edema scores or presence/absence of edema and increased expression of VEGF for any tumor group individually, or for all tumor groups combined.

Information relating to administration of corticosteroids was available for 35/37 MRI studies. Corticosteroids were administered in 12/37 dogs prior to MRI. There was no statistical association between the presence or absence of edema and corticosteroid administration.

Discussion

VEGF and its receptors are essential regulators of angiogenesis and vascular permeability and are postulated to play a major role in human tumor angiogenesis, carcinogenesis, invasion, and metastasis.⁴⁸ In the current study, we have demonstrated the presence of all 3 major VEGF mRNA isoforms in canine meningiomas, astrocytomas, and oligodendrogliomas and shown that VEGF mRNA expression is highest in high grade gliomas, specifically GBMs and grade III oligodendrogliomas.

Although 9 alternatively spliced human VEGF isoforms have been described (Fig. 1), 3 isoforms (VEGF₁₂₁, VEGF₁₆₅, and VEGF₁₈₉) predominate in the majority of human tissues and tumors. In particular, VEGF₁₆₅ and to a lesser extent VEGF₁₂₁ have been shown to be the predominant isoforms expressed in several human tumors, including astrocytomas, oligodendrogliomas, and meningiomas.^3, ^14, ^32, ³⁶ Individual isoforms have been shown to have distinct roles in vascular patterning and venous and arterial development; however, the roles of specific isoforms in neoplastic tissues remain unclear. The findings of the current study suggest that VEGF₁₆₄ (canine VEGF isoforms are 1 aa residue shorter than their human counterparts) is the predominant mRNA isoform in canine CNS tumors, similar to their human tumor counterparts, at least when significant overexpression is present. This predominance also parallels a previous study where VEGF₁₆₄ was found to be the major isoform in the canine heart.¹⁵ The TaqMan PCR primers designed to amplify VEGF₁₂₀, VEGF₁₆₄, and VEGF₁₈₈ for this study could potentially have amplified other additional isoforms based on human data (Fig. 1). Specifically, the VEGF₁₆₄ primers might also amplify VEGF₁₄₇ and VEGF_164b, and the VEGF₁₈₈ primers might amplify VEGF₁₄₅, VEGF₁₆₂, VEGF₁₈₂, and VEGF₂₀₅. The VEGF₁₂₀ primers should not amplify any other published human isoforms. Of these isoforms, only VEGF₁₈₂ has been described in dogs;¹⁵ however, it is likely that other isoforms are also present and remain to be documented. Isoforms other than VEGF₁₂₁, VEGF₁₆₅, and VEGF₁₈₉ have only been described in relatively small amounts in humans. Agarose gel analysis and sequencing of canine products generated with the current primer sets revealed single bands with product sequences consistent with the expected major VEGF isoforms, making the presence of significant amounts of most of the potential additional isoforms unlikely.

Increased expression of VEGF mRNA in canine high grade gliomas is not surprising since the presence of microvascular proliferation, presumably driven in part by VEGF, is a histologic criterion for their classification. Similar findings have been reported in human studies with high grade gliomas having greater expression than meningiomas,^13, ^14, ⁴⁹ and GBMs and high grade (III) oligodendrogliomas having greater expression than low grade astrocytomas and low grade (II) oligodendrogliomas.⁵ Some studies, however, reported no marked differences in VEGF expression between different tumor types and grades.⁴⁵ Only 2/20 oligodendrogliomas in this study were low grade, and statistical analysis was not possible; however, VEGF mRNA expression for these samples were the second and fifth lowest values for the oligodendroglial tumors. Care should be taken when attempting to interpret individual samples since transcription levels were assessed for small samples of the whole tumor, and VEGF mRNA expression may vary markedly in different regions of a tumor depending on intratumoral factors such as necrosis and hypoxia. As such, significant variation between individual tumors is to be expected, as seen in this study, and overall conclusions should be restricted to describing broad trends in tumor types and grades as a whole. Although the current study demonstrates increased VEGF mRNA expression in high grade canine tumors, no specific comments can be made relating to VEGF protein expression. However, increased levels of both VEGF mRNA and protein have been documented independently in a variety of human CNS tumors,^6, ^7, ^13, ^14, ^19, ^21, ^29, ^33, ^39, ^45, ^49, ^52, ⁵⁴ and some studies have specifically demonstrated elevated VEGF protein associated with elevated VEGF mRNA in the same tumor samples.^36, ³⁸ It is therefore likely that the general patterns seen in VEGF mRNA expression in canine CNS tumors will be reflected in similar studies quantifying protein expression.

The effect of corticosteroid administration on VEGF expression is poorly defined. Corticosteroids have been shown to decrease VEGF expression, predominantly at the transcriptional level in vitro;⁸ however, in vivo, human tumors with peritumoral edema may express VEGF whether or not pretreated with corticosteroids.^27, ^47, ⁵¹ No significant difference was seen within individual tumor groups when comparing levels of VEGF mRNA expression in corticosteroid treated to untreated samples in the current study. However, decreased total VEGF expression with corticosteroid administration approached significance (P = .055) when looking at all tumors combined. Dosage and duration of corticosteroid administration varied markedly (from less than 24 hours to several weeks) between individual dogs, and it is possible that VEGF mRNA expression could have been decreased following treatment in certain tumors. However it is unlikely that this had a significant effect on the broad conclusions made regarding VEGF mRNA expression in the different canine tumors.

VEGF is known to be a potent factor resulting in increased vascular permeability and has been associated with the presence of peritumoral vasogenic edema in human CNS tumors. No significant association between the level of VEGF mRNA expression and the degree or presence of peritumoral edema was seen in this study for any group of tumors, although MRI data were only available for a subset of the tumor samples and overall sample numbers were low. Several human studies have reported a correlation between VEGF expression and peritumoral edema⁴⁷ or vascular permeability²⁶ in human gliomas, and the decreased edema following treatment with specific inhibitors of VEGFR tyrosine kinase receptors in human clinical trials is strong circumstantial evidence for a major role of VEGF in edemagenesis.¹ However, a lack of correlation also has been reported.⁵¹ A more consistent correlation between VEGF expression and edema has been reported for human meningiomas,^12, ^16, ^34, ^35, ^37, ⁴¹ although several other factors have been implicated, including tumor size and location, presence of a cerebral pial blood supply,^4, ⁵³ other cytokines, vascular obstruction, and intrinsic permeability differences in tumor blood vessels. None of these additional factors were assessed in the current study. As discussed above, the exact role of corticosteroid administration on VEGF expression and its subsequent effects on peritumoral edema in vivo are unclear. Although the limited data in the current report must be interpreted with caution, no significant differences were seen between corticosteroid treated and untreated tumors and the presence or degree of edema, in either canine meningiomas or gliomas, and it is likely that the presence of peritumoral edema, similar to human tumors, is dependent on many factors in addition to VEGF expression.

Both presence^31, ^33, ⁵⁴ and absence^19, ⁵⁰ of a correlation between VEGF expression and clinical outcome has been reported for individual human patients with specific types and grades of CNS tumors, and high VEGF expression has been associated with poor outcome in canine grade I meningiomas.⁴⁰ No attempt was made to correlate prognosis with VEGF mRNA expression in the current study; however, intuitively, the higher level of expression in high grade gliomas as a group is likely to be associated with a poorer prognosis relative to the lower expression in lower grade gliomas and meningiomas.

VEGF, with its multiple isoforms, each potentially having multiple effects, is one factor in a family of proteins involved in the highly complex processes of angiogenesis, carcinogenesis, and control of vascular permeability in CNS tumors. We have previously shown that mRNAs for the 2 major tyrosine kinase receptors for VEGF (VEGFR-1, VEGFR-2) are overexpressed in many canine CNS tumors, particularly high grade gliomas.⁹ The current study further supports a major role for VEGF in the biology of these tumors and reinforces the striking similarities between canine CNS tumors and their human counterparts. Although the validity and efficacy of VEGF targeted therapy in neuro-oncology has still to be proven, these data support the further investigation of VEGF expression as a prognostic marker and therapeutic target in canine primary CNS tumors.

Footnotes

Acknowledgements

This research was supported by the Paul C. and Borghild T. Petersen Foundation and the Center for Companion Animal Health, University of California, Davis, CA. We are grateful to John Doval for his assistance with the illustrations.

References

1 Batchelor

Sorensen

di Tomaso

Zhang

Duda

Cohen

Kozak

Cahill

Chen

Zhu

Ancukiewicz

Mrugala

Plotkin

Drappatz

Louis

Ivy

Scadden

Benner

Loeffler

Wen

Jain

: AZD2171, a pan-VEGF receptor tyrosine kinase inhibitor, normalizes tumor vasculature and alleviates edema in glioblastoma patients. Cancer Cell 11: 83–95, 2007

2 Bates

Cui

Doughty

Winkler

Sugiono

Shields

Peat

Gillatt

Harper

: VEGF165b, an inhibitory splice variant of vascular endothelial growth factor, is down-regulated in renal cell carcinoma. Cancer Res 62: 4123–4131, 2002

3 Berkman

Merrill

Reinhold

Monacci

Saxena

Clark

Robertson

Ali

Oldfield

: Expression of the vascular permeability factor/vascular endothelial growth factor gene in central nervous system neoplasms. J Clin Invest 91: 153–159, 1993

4 Bitzer

Opitz

Popp

Morgalla

Gruber

Heiss

Voigt

: Angiogenesis and brain oedema in intracranial meningiomas: influence of vascular endothelial growth factor. Acta Neurochir 140: 333–340, 1998

5 Chan

Leung

Wong

Yuen

Cheung

Fan

Chung

: Expression of vascular endothelial growth factor and its receptors in the anaplastic progression of astrocytoma, oligodendroglioma, and ependymoma. Am J Surg Pathol 22: 816–826, 1998

6 Christov

Adle-Biassette

Le Guerinel

Natchev

Gherardi

: Immunohistochemical detection of vascular endothelial growth factor (VEGF) in the vasculature of oligodendrogliomas. Neuropathol Appl Neurobiol 24: 29–35, 1998

7 Christov

Gherardi

: Vascular endothelial growth factor (VEGF) likely contributes to oligodendroglioma angiogenesis. Acta Neuropathol (Berl) 97: 429–432, 1999

8 Criscuolo

Merrill

Oldfield

: Further characterization of malignant glioma-derived vascular permeability factor. J Neurosurg 69: 254–262, 1988

9 Dickinson

Roberts

Higgins

Leutenegger

Bollen

Kass

LeCouteur

: Expression of receptor tyrosine kinases VEGFR-1 (FLT-1), VEGFR-2 (KDR), EGFR-1, PDGFRa and c-Met in canine primary brain tumors. Veterinary and Comparative Oncology 4: 132–140, 2006

10.

10 Ferrara

Davis-Smyth

: The biology of vascular endothelial growth factor. Endocr Rev 18: 4–25, 1997

11.

11 Ferrara

Gerber

LeCouter

: The biology of VEGF and its receptors. Nat Med 9: 669–676, 2003

12.

12 Goldman

Bharara

Palmer

Vitek

Tsai

Weiss

Gillespie

: Brain edema in meningiomas is associated with increased vascular endothelial growth factor expression. Neurosurgery 40: 1269–1277, 1997

13.

13 Hatva

Kaipainen

Mentula

Jaaskelainen

Paetau

Haltia

Alitalo

: Expression of endothelial cell-specific receptor tyrosine kinases and growth factors in human brain tumors. Am J Pathol 146: 368–378, 1995

14.

14 Huang

Held-Feindt

Buhl

Mehdorn

Mentlein

: Expression of VEGF and its receptors in different brain tumors. Neurol Res 27: 371–377, 2005

15.

15 Jingjing

Srinivasan

Bian

Downey

Roque

: Vascular endothelial growth factor is increased following coronary artery occlusion in the dog heart. Mol Cell Biochem 214: 23–30, 2000

16.

16 Kalkanis

Carroll

Zhang

Zamani

Black

: Correlation of vascular endothelial growth factor messenger RNA expression with peritumoral vasogenic cerebral edema in meningiomas. J Neurosurg 85: 1095–1101, 1996

17.

17 Keck

Hauser

Krivi

Sanzo

Warren

Feder

Connolly

: Vascular permeability factor, an endothelial cell mitogen related to PDGF. Science 246: 1309–1312, 1989

18.

18 Kleihues

Louis

Scheithauer

Rorke

Reifenberger

Burger

Cavenee

: The WHO classification of tumors of the nervous system. J Neuropathol Exp Neurol 61: 215–225; discussion 226–229, 2002

19.

19 Korshunov

Golanov

: The prognostic significance of vascular endothelial growth factor (VEGF C-1) immunoexpression in oligodendroglioma. An analysis of 91 cases. J Neurooncol 48: 13–19, 2000

20.

20 Lafuente

Adan

Alkiza

Garibi

Rossi

Cruz-Sanchez

: Expression of vascular endothelial growth factor (VEGF) and platelet-derived growth factor receptor-beta (PDGFR-beta) in human gliomas. J Mol Neurosci 13: 177–185, 1999

21.

21 Lamszus

Lengler

Schmidt

Stavrou

Ergun

Westphal

: Vascular endothelial growth factor, hepatocyte growth factor/scatter factor, basic fibroblast growth factor, and placenta growth factor in human meningiomas and their relation to angiogenesis and malignancy. Neurosurgery 46: 938–947; discussion 947–938, 2000

22.

22 Leung

Cachianes

Kuang

Goeddel

Ferrara

: Vascular endothelial growth factor is a secreted angiogenic mitogen. Science 246: 1306–1309, 1989

23.

23 Lipsitz

Higgins

Kortz

Dickinson

Bollen

LeCouteur

: Glioblastoma multiforme: clinical findings, magnetic resonance imaging and pathology in 5 dogs. Vet Pathol 40: 659–669, 2003

24.

24 Louis

Ohgaki

Wiestler

Cavenee

: WHO classification of tumors of the central nervous system, 4th ed, WHO Press, Geneva, Switzerland, 2007

25.

25 Louis

Ohgaki

Wiestler

Cavenee

Burger

Jouvet

Scheithauer

Kleihues

: The 2007 WHO Classification of Tumours of the Central Nervous System. Acta Neuropathol (Berl) 114: 97–109, 2007

26.

26 Machein

Kullmer

Fiebich

Plate

Warnke

: Vascular endothelial growth factor expression, vascular volume, and, capillary permeability in human brain tumors. Neurosurgery 44: 732–740; discussion 740–731, 1999

27.

27 Machein

Kullmer

Ronicke

Machein

Krieg

Damert

Breier

Risau

Plate

: Differential downregulation of vascular endothelial growth factor by dexamethasone in normoxic and hypoxic rat glioma cells. Neuropathol Appl Neurobiol 25: 104–112, 1999

28.

28 Machein

Plate

: VEGF in brain tumors. J Neurooncol 50: 109–120, 2000

29.

29 Maia

Jr Malheiros

da Rocha

da Silva

Gabbai

Ferraz

Stavale

: MR cerebral blood volume maps correlated with vascular endothelial growth factor expression and tumor grade in nonenhancing gliomas. AJNR Am J Neuroradiol 26: 777–783, 2005

30.

30 Maiolino

De Vico

Restucci

: Expression of vascular endothelial growth factor in basal cell tumours and in squamous cell carcinomas of canine skin. J Comp Pathol 123: 141–145, 2000

31.

31 Maiuri

Del Basso De Caro

Iaconetta

Peca

Esposito

de Divitiis

: Prognostic and survival-related factors in patients with well-differentiated oligodendrogliomas. Zentralbl Neurochir 67: 204–209, 2006

32.

32 Munaut

Boniver

Foidart

Deprez

: Macrophage migration inhibitory factor (MIF) expression in human glioblastomas correlates with vascular endothelial growth factor (VEGF) expression. Neuropathol Appl Neurobiol 28: 452–460, 2002

33.

33 Nam

Park

Suh

Kim

: Expression of VEGF and brain specific angiogenesis inhibitor-1 in glioblastoma: prognostic significance. Oncol Rep 11: 863–869, 2004

34.

34 Otsuka

Tamiya

Ono

Michiue

Kurozumi

Daido

Kambara

Date

Ohmoto

: The relationship between peritumoral brain edema and the expression of vascular endothelial growth factor and its receptors in intracranial meningiomas. J Neurooncol 70: 349–357, 2004

35.

35 Paek

Kim

Park

Kim

Jung

: Correlation of clinical and biological parameters with peritumoral edema in meningioma. J Neurooncol 60: 235–245, 2002

36.

36 Pistolesi

Boldrini

Gisfredi

De Ieso

Camacci

Caniglia

Lupi

Leocata

Basolo

Pingitore

Parenti

Fontanini

: Angiogenesis in intracranial meningiomas: immunohistochemical and molecular study. Neuropathol Appl Neurobiol 30: 118–125, 2004

37.

37 Pistolesi

Fontanini

Camacci

De Ieso

Boldrini

Lupi

Padolecchia

Pingitore

Parenti

: Meningioma-associated brain oedema: the role of angiogenic factors and pial blood supply. J Neurooncol 60: 159–164, 2002

38.

38 Plate

Breier

Weich

Mennel

Risau

: Vascular endothelial growth factor and glioma angiogenesis: coordinate induction of VEGF receptors, distribution of VEGF protein and possible in vivo regulatory mechanisms. Int J Cancer 59: 520–529, 1994

39.

39 Plate

Breier

Weich

Risau

: Vascular endothelial growth factor is a potential tumour angiogenesis factor in human gliomas in vivo. Nature 359: 845–848, 1992

40.

40 Platt

Scase

Adams

Wieczorek

Miller

Adamo

Long

: Vascular endothelial growth factor expression in canine intracranial meningiomas and association with patient survival. J Vet Intern Med 20: 663–668, 2006

41.

41 Provias

Claffey

delAguila

Lau

Feldkamp

Guha

: Meningiomas: role of vascular endothelial growth factor/vascular permeability factor in angiogenesis and peritumoral edema. Neurosurgery 40: 1016–1026, 1997

42.

42 Restucci

Maiolino

Paciello

Martano

De Vico

Papparella

: Evaluation of angiogenesis in canine seminomas by quantitative immunohistochemistry. J Comp Pathol 128: 252–259, 2003

43.

43 Restucci

Papparella

Maiolino

De Vico

: Expression of vascular endothelial growth factor in canine mammary tumors. Vet Pathol 39: 488–493, 2002

44.

44 Robinson

Stringer

: The splice variants of vascular endothelial growth factor (VEGF) and their receptors. J Cell Sci 114: 853–865, 2001

45.

45 Samoto

Ikezaki

Ono

Shono

Kohno

Kuwano

Fukui

: Expression of vascular endothelial growth factor and its possible relation with neovascularization in human brain tumors. Cancer Res 55: 1189–1193, 1995

46.

46 Scheidegger

Weiglhofer

Suarez

Kaser-Hotz

Steiner

Ballmer-Hofer

Jaussi

: Vascular endothelial growth factor (VEGF) and its receptors in tumor-bearing dogs. Biol Chem 380: 1449–1454, 1999

47.

47 Strugar

Criscuolo

Rothbart

Harrington

: Vascular endothelial growth/permeability factor expression in human glioma specimens: correlation with vasogenic brain edema and tumor-associated cysts. J Neurosurg 83: 682–689, 1995

48.

48 Takahashi

Shibuya

: The vascular endothelial growth factor (VEGF)/VEGF receptor system and its role under physiological and pathological conditions. Clin Sci (Lond) 109: 227–241, 2005

49.

49 Takano

Yoshii

Kondo

Suzuki

Maruno

Shirai

Nose

: Concentration of vascular endothelial growth factor in the serum and tumor tissue of brain tumor patients. Cancer Res 56: 2185–2190, 1996

50.

50 Vaquero

Zurita

Coca

Oya

Morales

: Prognostic significance of clinical and angiogenesis-related factors in low-grade oligodendrogliomas. Surg Neurol 54: 229–234; discussion 234, 2000

51.

51 Vaquero

Zurita

Morales

Cincu

Oya

: Expression of vascular permeability factor in glioblastoma specimens: correlation with tumor vascular endothelial surface and peritumoral edema. J Neurooncol 49: 49–55, 2000

52.

52 Varlet

Guillamo

Nataf

Koziak

Beuvon

Daumas-Duport

: Vascular endothelial growth factor expression in oligodendrogliomas: a correlative study with Sainte-Anne malignancy grade, growth fraction and patient survival. Neuropathol Appl Neurobiol 26: 379–389, 2000

53.

53 Yoshioka

Hama

Taniguchi

Sugiyama

Arita

Kurisu

: Peritumoral brain edema associated with meningioma: influence of vascular endothelial growth factor expression and vascular blood supply. Cancer 85: 936–944, 1999

54.

54 Zhou

Tan

Hess

Yung

: The expression of PAX6, PTEN, vascular endothelial growth factor, and epidermal growth factor receptor in gliomas: relationship to tumor grade and survival. Clin Cancer Res 9: 3369–3375, 2003