The delta-like ligand 4-notch signaling inhibits tumor angiogenesis but promotes tumor growth in primary glioblastoma: An immunohistochemical study

Yao Chen; Zhi-Xiong Lin; Jin-Tao Chen; Ming-Cheng Zheng

doi:10.4103/glioma.glioma_11_18

ORIGINAL ARTICLE

Year : 2018 | Volume : 1 | Issue : 2 | Page : 59-65

The delta-like ligand 4-notch signaling inhibits tumor angiogenesis but promotes tumor growth in primary glioblastoma: An immunohistochemical study

Yao Chen¹, Zhi-Xiong Lin², Jin-Tao Chen³, Ming-Cheng Zheng³
¹ Department of Neurosurgery, The Affiliated Hospital of Putian University, Putian, Fujian, China
² Department of Neurosurgery, Beijing Sanbo Brain Hospital, Capital Medical University, Beijing, China
³ Tumor Invasion Micro-Ecological Laboratory, Fujian Medical University, Fuzhou, Fujian, China

Date of Web Publication

30-Apr-2018

Correspondence Address:
Dr. Zhi-Xiong Lin
Department of Neurosurgery, Beijing Sanbo Brain Hospital, Capital Medical University, Beijing 100093
China

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/glioma.glioma_11_18

Abstract

Background: Delta-like ligand 4 (DLL4) is a key Notch ligand implicated in tumor angiogenesis. However, previous studies have shown that the DLL4-Notch signaling inhibits tumor angiogenesis but promotes tumor progression in primary glioblastoma. The underlying mechanism remains unknown.
Methods: Tumor tissues from 70 patients with primary glioblastoma were analyzed by immunohistochemistry for the expression of DLL4, microvessel density (MVD), and Ki67 labeling index. The degree of tumor contrast enhancement (DTCE) on preoperative magnetic resonance imaging was also evaluated. The effect on prognosis was assessed based on Kaplan–Meier survival and Cox proportional hazard models.
Results: Results showed that the DTCE was negatively correlated with DLL4 but positively correlated with MVD (r = −0.260, 0.593, P < 0.05). The Ki67 labeling index was shown to be positively correlated with both DLL4 and MVD (r = 0.346, 0.346, P < 0.05). Univariate analysis indicated a significant correlation of high DLL4 and Ki67 labeling index expression with shorter progression-free survival (PFS) and overall survival (OS) (P < 0.05). Multivariate analysis confirmed high DLL4 and MVD expression as unfavorable prognostic indicators for PFS and OS (P < 0.05), and the hazard ratio of DLL4 was higher than MVD for both PFS and OS (P < 0.05).
Conclusion: We conclude that the DLL4-Notch signaling improves tumor vascular function and contributes to the survival of malignant cells, resulting in less MVD but more tumor progression in primary glioblastoma.

Keywords: Angiogenesis, delta-like ligand 4-notch signaling, microvessel density, primary glioblastoma, prognosis

How to cite this article:
Chen Y, Lin ZX, Chen JT, Zheng MC. The delta-like ligand 4-notch signaling inhibits tumor angiogenesis but promotes tumor growth in primary glioblastoma: An immunohistochemical study. Glioma 2018;1:59-65

How to cite this URL:
Chen Y, Lin ZX, Chen JT, Zheng MC. The delta-like ligand 4-notch signaling inhibits tumor angiogenesis but promotes tumor growth in primary glioblastoma: An immunohistochemical study. Glioma [serial online] 2018 [cited 2023 Jun 4];1:59-65. Available from: http://www.jglioma.com/text.asp?2018/1/2/59/231496

Yao Chen and Zhi-Xiong Lin contributed equally to this work.

Introduction

Glioblastoma is the most common primary malignant brain tumor in adults, accounting for >50% of primary malignant gliomas.^[1] After maximum surgical tumor resection, standard therapy is a concurrent treatment of temozolomide (TMZ) and radiotherapy.^[2] However, the clinical outcome of primary glioblastoma remains poor, with a median survival of 18 months.^[3] Therefore, new therapeutic strategies are required.

In the 1970s, Folkman ^[4] first proposed the concept that solid tumors require the growth of new blood vessels for oxygen and nutrient supply, and the antiangiogenesis therapy based on this concept has spurred substantial efforts in both preclinical and clinical studies. Among tumor angiogenesis, vascular endothelial growth factor (VEGF) signaling is the most important and best characterized pathway.^[5] Studies have found that blocking the VEGF signaling pathway can inhibit the growth of tumor cells, as shown in various mouse tumor models.^[6] Thus, studies hope to improve the prognosis of glioblastoma by antiangiogenesis. However, two phase III clinical trials in 2014 have shown that blocking the VEGF signaling pathway through bevacizumab did not improve overall survival (OS) in patients with newly diagnosed glioblastoma.^[7],[8] Therefore, it is of great significance to study the role of other signaling pathways in the formation of tumor angiogenesis in human glioblastoma.

The Notch pathway is a highly conserved signaling system that controls a diversity of growth, differentiation, and pattern processes throughout embryonic and postnatal development in vertebrates and invertebrates.^[9],[10] Until recently, 4 types of Notch receptors: Notch-1, -2, -3, and -4 and five ligands: Delta-like-1, -3, -4 (DLL1, 3, 4), Jagged-1 and -2 have been recognized in mammals.^[10],[11] Accumulating evidence shows that the DLL4-Notch pathway can enhance the activity of VEGF on tumor cells, promote its expression, and stimulate the formation of tumor angiogenesis to promote tumor growth/invasion.^{[12],[13],[14]} Studies of tumor angiogenesis in physiological vessels show that upregulated DLL4 causes the activation of Notch signaling in the vascular endothelial cells.^[15]In vitro, hypoxia generally induces the abnormal expression of DLL4, which in turn increases the microvessel density (MVD).^[16] However, the function of DLL4-Notch signaling pathway in the formation of tumor angiogenesis in glioblastoma is not uniform. In most of previous studies, the DLL4-Notch signaling pathway inhibited the formation of tumor angiogenesis in glioblastoma,^[17],[18] but an individual study shows the opposite result.^[19] In our previous study, we also found that high DLL4 expression predicted poor prognosis in primary glioblastoma.^[20] If this result is consistent with the majority of studies (the DLL4-Notch signaling pathway inhibited glioblastoma tumor angiogenesis), it is seemingly contradictory with the traditional concept: "Antitumor angiogenesis can serve as a therapeutic target for cancer." For this reason, we have done mainly two things. One was to clarify the role of DLL4-Notch signaling pathway in the tumor angiogenesis in glioblastoma. More importantly, we explored the different effects of DLL4-Notch signaling pathway on the prognosis of patients with glioblastoma to explain the possible reasons and mechanism of this contradictory result. To the best of our knowledge, there is currently a lack of research on this aspect.

Materials and Methods

Patients and tissue samples

This study retrospectively identified patients with primary glioblastoma who underwent surgery at the Department of Neurosurgery, the First Affiliated Hospital of Fujian Medical University between 2009 and 2014. Patients with recurrent glioblastoma through simple biopsy, without preoperative magnetic resonance imaging (MRI) image, as well as patients whose enhanced scan image data and postoperative pathologic specimens were not available for immunohistochemically were excluded. In total, 70 patients were enrolled. Informed consent was obtained from all patients according to the research proposal approved by the local ethics committee of the Fujian Medical University. All procedures performed in this study were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki declaration and its later amendments. Tumor specimens were reviewed and confirmed by two experienced neuropathologists who were blinded to the study, using the principles of WHO classification for tumors in the central nervous system.^[1] None of the patients received preoperative adjuvant therapy (radiotherapy and/or chemotherapy). Clinical information was obtained by reviewing medical records.

Treatment and clinical outcome assessment

Following diagnosis, all patients underwent surgical treatment. According to the postoperative MRI performed within 7 days after surgery, the extent of tumor resection was determined as "gross total resection" (≥95% of tumor was resected) in 55 patients and as "subtotal resection" (<95% resected) in 15 patients. Following surgery, radiotherapy to limited fields (2 Gy per fraction, once a day, 5 days a week, 60 Gy total dose) plus continuous daily TMZ (75 mg/m ²/d, once per day, 7 times/week during radiotherapy) was adopted. Subsequently, adjuvant TMZ (at a dose of 150–200 mg/m ²/d, days 1–5) was used in chemotherapy and administered for 4–6 cycles (at 28-day intervals) in the absence of irreversible blood toxicity or death. Nineteen patients were treated with nonstandard chemoradiotherapy (including patients received only radiotherapy or chemotherapy and patients did not receive chemoradiotherapy) and 51 patients received standard chemoradiotherapy (radiotherapy plus chemotherapy). All patients in this study were followed up in accordance with a strict protocol as follows. After the date of resection, patients were observed at 3-month intervals during the 1^st year and at 6-month intervals thereafter. Tumor progression was determined based on the method proposed by Wen et al.^[21] Progression-free survival (PFS) was measured from the date of surgery until the date of tumor progression (or the latest follow-up). OS was measured from the time of surgery to death (or the latest follow-up). PFS and OS were censored at the date of latest follow-up for patients without tumor progression or death. The clinical characteristics of patient data are presented in [Table 1].

Table 1: Clinical characteristics of patient

Click here to view

Immunohistochemistry staining

Sections were cut at a thickness of 4 μm from routinely prepared, 10% formalin-fixed, paraffin-embedded blocks of tumor tissue. The following primary antibodies were used: anti-DLL4 antibody (1:100, Abcam, ab7280, UK), anti-CD34 antibody (1:50, Dako, clone QBEnd/10, Denmark), and anti-Ki67 antibody (1:100, Dako, clone Mib1, Denmark). Paraffin-embedded sections were deparaffinized in xylene and dehydrated in graded alcohol. Antigen retrieval was performed in a citrate buffer (pH 6.0). The sections were incubated overnight at 4°C with the diluted primary antibodies. Next, the sections were rinsed with phosphate-buffered solution (PBS) and incubated with the horseradish peroxidase-conjugated secondary antibody (ZSGB-BIO, China). After rinsing with PBS, the sections were stained with diaminobenzidine and counterstained with hematoxylin. The negative control sections were incubated with PBS in equal concentrations as the primary antibodies, and known positive tissue sections (human kidney tissue for DLL4; human lung carcinoma tissue for CD34) were used as positive controls.

Evaluation of staining

Two pathologists, who were blinded to the pathological diagnosis and clinical data, analyzed the immunohistochemical staining. The number of positive endothelial or tumor cells was counted using a light microscope at ×200. Five fields were examined for each tumor specimen. The expression of DLL4 was classified as low if ≤10% of the cells were positive; otherwise was classified as high.^[22] For the Ki67 labeling index, a semiquantitative grading scale was adopted as follows: (1) weak immunoreactivity (5%–25%), (2) moderate immunoreactivity (25%–50%), and (3) strong immunoreactivity (≥50%). To facilitate the analysis, we assumed 0-1 as low, and others as high.

Determination of microvessel density

The MVD was determined by counting CD34-labeled microvessel endothelial cells. Due to the uneven distribution of microvessels in tumor tissue, each section displaying dual staining of CD34 was scanned at a low magnification (×40) and three regions with the highest density of distinctly highlighted microvessels ("hot spots") were selected. One field was identified in each hot spot at ×200 (0.785 mm ² per field size). MVD was defined as the number of manually counted vessels per mm ² and presented as the mean of three hot spots.^[23] The median of MVD served as the cutoff: the expression was classified as low if less than the median, otherwise was classified as high.

Evaluation of preoperative magnetic resonance imaging

For all patients, preoperative MRI data were acquired from available routine scans, including T1-weighted (W), T2-W, and contrast-enhanced T1-W sequences. Tumor contrast enhancement was defined as contrast-enhanced T1-W sequences compared with T1-W display of the high signal [Figure 1]. The degree of tumor contrast enhancement (DTCE) was measured on postcontrast T1-W and divided into 3 grades:^[24] 0 = no enhancement; 1 = signal intensity of the tumor nodule was less than that of fat; and 2 = signal intensity of the tumor nodule was equal to or higher than that of fat. Since all patients in this study showed enhancement on postcontrast T1-W, Degree 1 was defined as low enhancement [Figure 1]B and Degree 2 was defined as high enhancement [Figure 1]D. Each patient's MRI images were analyzed independently by two experienced radiologists who were blinded to the patient's clinical information.

Figure 1: The degree of tumor contrast enhancement of patients with primary glioblastoma. (A and B) axial T1-weighted (w) images of the same patient: (A) axial precontrast T1-W image; (B) axial contrast-enhanced T1-W image showed low enhancement. (C and D) axial T1-weighted (w) images of another patient: (C) axial precontrast T1-W image; (D) axial contrast-enhanced T1-W image showed high enhancement

Click here to view

Statistical analysis

All data were analyzed using SPSS 22.0 software (SPSS Inc., Chicago, USA). Relationships between categorical variables (age and gender) and DLL4, Ki67, MVD, and DTCE were based on the Chi-square test. A nonparametric test (Spearman's rank correlation (r) analysis or Mann–Whitney U-test) was adopted to analyze the association between DLL, MVD, Ki67, and DTCE. Survival curves were constructed based on the Kaplan–Meier method, and survival differences were compared based on a log-rank test. A multivariate analysis was done by using Cox proportional hazards model to determine the prognostic effectors on PFS and OS. Hazard ratios (HR) and corresponding 95% confidence intervals (CI) were calculated. P ≤ 0.05 was considered to be statistically significant.

Results

Topology of delta-like ligand 4, Ki67, CD34 (microvessel density) in primary glioblastoma tissues

As compared to adjacent nonneoplastic brain tissue with low or rare immunoreactivity, DLL4, Ki67, and CD34 exhibited elevated expression. Positive staining of DLL4 expression was essentially limited to the cytoplasm of tumor vascular endothelial cells rather than tumor cells [Figure 2]A; Ki67 immunostaining was detected in the nucleus of both neoplastic cells and tumor vascular endothelial cells [Figure 2]B.

Figure 2: Immunohistochemical staining for delta-like ligand 4, Ki67, and CD34 expression. (A and B) Positive staining of delta-like ligand 4, Ki67 in primary glioblastoma (×400): (A) positive staining of delta-like ligand 4 primarily distributed in the cytoplasm of tumor endothelial cells and seldom detected in some tumor cells; (B) positive nuclear staining of Ki67 in both tumor endothelial cells and tumor cells. Positive CD34 staining in primary glioblastoma (C) and adjacent nonneoplastic brain tissues (D) (×200): (C) CD34 (microvessel density) high expressed in primary glioblastoma; (D) low CD34 (microvessel density) expression was seen in adjacent nonneoplastic brain tissues (positive staining in tumor endothelial cells: red arrow; positive staining in tumor cells: green arrow)

Click here to view

In general, MVD is a well-known parameter associated with angiogenesis in tumor. In the present study, MVD was evaluated in primary glioblastoma paraffin sections after staining with CD34. CD34 expression was observed in the cytoplasm of tumor vascular endothelial cells and the microvascular formation varied, including single endothelium and tubular vessels in primary glioblastoma tissues [Figure 2]D. The mean of MVD in primary glioblastoma tissues was 97.07 ± 35.17 mm ² (range 30.57–168.15 mm ²), which was significantly higher than in nonneoplastic brain tissues (58.91 ± 21.80 mm ², range 30.57–87.90 mm ²) (P = 0.001) [Figure 2]C and [Figure 2]D and [Figure 3]B.

Figure 3: Expression of microvessel density. (A) Microvessel density in low delta-like ligand 4 expression group is higher than that in high delta-like ligand 4 expression group. (B) Microvessel density in primary glioblastoma tissues is higher than in adjacent nonneoplastic brain tissues

Click here to view

Correlations of delta-like ligand 4, Ki67, microvessel density, and degree of tumor contrast enhancement with clinical features

The total concordance of scoring between the 2 observers was 95.7% for DLL4 (67/70, Cohen's kappa coefficient of 0.746, P < 0.001) and 92.8% for Ki67 (65/70, Cohen's kappa coefficient of 0.851, P < 0.001) indicating substantial agreement. Of 70 patients, 29 (41.4%) were females, 41 (58.6%) were males (1:1.4 sex ratio), 41 patients (58.6%) were ≤60 years, 29 (41.4%) were >60 years, and the mean age at diagnosis was 59.56 ± 12.05 years (17–79 years). Preoperatively, median karnofsky performance status (KPS) was 70 (range, 60–90) [Table 1]. Our results demonstrated that all 70 patients got enhanced on preoperative postcontrast MRI scans, among them 29 showed low degree enhancement, while 41 showed high degree enhancement. However, there was no correlation found with gender and age (P > 0.05) [Table 2].

Table 2: Correlations of delta-like ligand 4, Ki67, microvessel density, and enhancement with clinical variables

Click here to view

Relationship between delta-like ligand 4, Ki67, microvessel density, and degree of tumor contrast enhancement

Our results showed that DLL4 expression was significantly associated with MVD, DTCE, and Ki67 (P < 0.05) [Table 3]. The correlation analysis indicated that DLL4 expression was negatively correlated with MVD (r = −0.249, P = 0.038) [Figure 3]A and DTCE (r = −0.260, P = 0.030) but positively correlated with Ki67 (r = 0.346, P = 0.003) [Table 3]. Our results regarding tumor angiogenesis and progression showed that MVD was positive correlated with both DTCE and Ki67 (r = 0.593, 0.346, P < 0.05) [Table 3].

Table 3: Relationship between delta-like ligand 4, microvessel density, Ki67, and degree of tumor contrast enhancement

Click here to view

Survival analysis

Follow-up was available for all patients, and the mean follow-up period was 48.14 weeks (range: 11–154). During the follow-up period, 65 (92.9%) patients showed tumor progression and 45 (64.3%) patients died, and all of them had died with glioblastoma. Univariate analysis demonstrated that patients with low DLL4 and MVD expression had significantly longer PFS (P ≤ 0.001, 0.024) [Figure 4]A and [Figure 4]C and OS (P ≤ 0.001, 0.007) [Figure 4]B and [Figure 4]D than those with high expression. Similar results were also observed between survival estimates of younger patients, with higher degrees of tumor resection, standard chemoradiotherapy, higher KPS, and low Ki67 expression [Table 4].

Figure 4: Effect of microvessel density and delta-like ligand 4 on progression-free survival and overall survival in glioblastoma. (A and B) High microvessel density and (C and D) high delta-like ligand 4 are associated with both reduced progression-free survival and overall survival

Click here to view

Table 4: Univariate and multivariate analyses of variable associated with progression-free survival and overall survival

Click here to view

Multivariate analysis confirmed that low expression of DLL4 and MVD is a significant prognostic predictor for longer PFS (P = 0.005, 0.035) and OS (P = 0.004, 0.004). The HRs of DLL4 for PFS and OS (2.896 [95% CI 1.381–6.074], 3.208 [95% CI 1.465–7.026], respectively [P < 0.05]) were higher than MVD (2.509 [95% CI 1.278–4.926], 2.890 [95% CI 1.393–5.996], [P < 0.05]) [Table 4].

Discussion

The DLL4 is a key Notch ligand that is implicated in tumor angiogenesis. However, there are uncertainties regarding the function of the DLL4-Notch signaling pathway in the formation of tumor angiogenesis in glioblastoma.^{[17],[18],[19]} Although most studies in the literature have concluded that the DLL4-Notch signaling pathway is a negative effector on tumor angiogenesis, our study has shown that high DLL4 expression predicted poor prognosis in primary glioblastoma.^[20] Reasons and mechanisms of this contradictory result have not been reported.

The present study showed that MVD in primary glioblastoma was significantly higher than in nonneoplastic brain tissue, and the spatial distribution of DLL4 was mainly located in the tumor vascular endothelial cells. This significant association and special morphology suggests DLL4 as a negative effector on tumor angiogenesis in primary glioblastoma. The results also demonstrated that both DLL4 and MVD were associated with Ki67. In the present study, there was a positive correlation between the expression of MVD and Ki67. The possible reason may be that high MVD can provide more nutrients and oxygen for tumor progression.^[4] However, although this study showed that DLL4 inhibited MVD, a positive correlation between DLL4 and Ki67 was also found in primary glioblastoma. Therefore, the further study was done by us.

Tumor enhancement is one of the most common imaging features of primary glioblastoma resulted in part from the abundance of neovascularization. When considering MRI-enhanced images, normal brain tissue cannot be enhanced due to the protection of blood–brain barrier (BBB). However, in primary glioblastoma tissues, the abundance of neovascularization leads to the formation of abnormal blood vessels, with absent or incomplete BBB, which allows enhancement by the contrast agent.^[25] Therefore, the DTCE is a comprehensive reflection of the function of the tumor vessel and tumor angiogenesis. In the present study, we observed that the DTCE was positively correlated with MVD but negatively correlated with DLL4. This phenomenon may be related to the different functions with respect to tumor angiogenesis. An abundance of pathological angiogenesis with high permeability can create a powerful condition for the migration of tumor vascular endothelial cells and facilitates proliferation and invasion of tumor cells.^[13] Thus, when MRI enhanced scans were conducted, more contrast agent exuded from tumor tissue. In contrast, DLL4-Notch signaling pathway decreases tumor angiogenesis ^[26] but improves tumor vascular function in primary glioblastoma.^[27] Therefore, low DTCE was observed.

Furthermore, we evaluated the effect of DLL4 and MVD expression on patient outcomes in glioblastoma. These results showed that both the DLL4 and MVD were independent predictors of prognosis. High MVD can provide more nutrients and oxygen for tumor progression.^[4] However, although DLL4 reduces the formation of blood vessels, it improves tumor vascular function and produces "bigger" lumen within vessels to transport nutrients and oxygen for tumor growth.^[27] In addition, the DLL4-Notch signaling pathway nurtures the self-renewal of cancer stem-like cells,^[28] which leads to tumor invasion, proliferation, and metastasis. Moreover, from the COX proportional hazard model, we found that the HR of DLL4 for OS was higher than MVD (3.208 vs. 2.890). These findings indicate that in the process of tumor development, the function of the vessel was equally important and may be even more than its density. This may be one of the reasons why targeted therapy with VEGF antagonists alone did not work effectively in clinical studies.^[29],[30]

However, our study has several limitations which should be acknowledged. First and foremost, the findings are limited because of the few samples and the conclusions based mainly on the statistical analysis of immunohistochemistry results. Therefore, large-scale and vitro or vivo experiments are worth doing to provide more compelling evidence.

In conclusion, our results suggest that although the DLL4-Notch signaling pathway inhibits tumor angiogenesis in primary glioblastoma, its function also improves tumor vascular. In carcinogenesis, this may lead to poor outcomes. On the other hand, the function of the vessel was equally important and may potentially be more important than density. Our data also demonstrate that DLL4-Notch signaling (and its inhibitors) may be used as a promising therapeutic target in glioblastoma.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

References

1.	Komori T, Sasaki H, Yoshida K. Revised WHO classification of tumours of the central nervous system: Summary of the revision and perspective. No Shinkei Geka 2016;44:625-35. [PUBMED]
2.	Dubrow R, Darefsky AS, Jacobs DI, Park LS, Rose MG, Laurans MS, et al. Time trends in glioblastoma multiforme survival: The role of temozolomide. Neuro Oncol 2013;15:1750-61. [PUBMED]
3.	Cai X, Sughrue ME. Glioblastoma: New therapeutic strategies to address cellular and genomic complexity. Oncotarget 2018;9:9540-54. [PUBMED]
4.	Folkman J. Tumor angiogenesis: Therapeutic implications. N Engl J Med 1971;285:1182-6. [PUBMED]
5.	Carmeliet P. Angiogenesis in life, disease and medicine. Nature 2005;438:932-6. [PUBMED]
6.	Gerber HP, Ferrara N. Pharmacology and pharmacodynamics of bevacizumab as monotherapy or in combination with cytotoxic therapy in preclinical studies. Cancer Res 2005;65:671-80. [PUBMED]
7.	Gilbert MR, Dignam JJ, Armstrong TS, Wefel JS, Blumenthal DT, Vogelbaum MA, et al. A randomized trial of bevacizumab for newly diagnosed glioblastoma. N Engl J Med 2014;370:699-708. [PUBMED]
8.	Chinot OL, Wick W, Mason W, Henriksson R, Saran F, Nishikawa R, et al. Bevacizumab plus radiotherapy-temozolomide for newly diagnosed glioblastoma. N Engl J Med 2014;370:709-22. [PUBMED]
9.	Yamamoto S, Schulze KL, Bellen HJ. Introduction to Notch signaling. Methods Mol Biol 2014;1187:1-4. [PUBMED]
10.	Artavanis-Tsakonas S, Rand MD, Lake RJ. Notch signaling: Cell fate control and signal integration in development. Science 1999;284:770-6. [PUBMED]
11.	Fleming RJ. Structural conservation of Notch receptors and ligands. Semin Cell Dev Biol 1998;9:599-607. [PUBMED]
12.	Benedito R, Roca C, Sörensen I, Adams S, Gossler A, Fruttiger M, et al. The Notch ligands dll4 and Jagged1 have opposing effects on angiogenesis. Cell 2009;137:1124-35.
13.	Hernandez SL, Banerjee D, Garcia A, Kangsamaksin T, Cheng WY, Anastassiou D, et al. Notch and VEGF pathways play distinct but complementary roles in tumor angiogenesis. Vasc Cell 2013;5:17.
14.	Suchting S, Freitas C, le Noble F, Benedito R, Bréant C, Duarte A, et al. The notch ligand delta-like 4 negatively regulates endothelial tip cell formation and vessel branching. Proc Natl Acad Sci U S A 2007;104:3225-30.
15.	Pedrosa AR, Trindade A, Fernandes AC, Carvalho C, Gigante J, Tavares AT, et al. Endothelial Jagged1 antagonizes dll4 regulation of endothelial branching and promotes vascular maturation downstream of dll4/Notch1. Arterioscler Thromb Vasc Biol 2015;35:1134-46. [PUBMED]
16.	Stockhausen MT, Kristoffersen K, Poulsen HS. The functional role of Notch signaling in human gliomas. Neuro Oncol 2010;12:199-211. [PUBMED]
17.	Zhang JF, Chen Y, Qiu XX, Tang WL, Zhang JD, Huang JH, et al. The vascular delta-like ligand-4 (DLL4)-Notch4 signaling correlates with angiogenesis in primary glioblastoma: An immunohistochemical study. Tumour Biol 2016;37:3797-805. [PUBMED]
18.	El Hindy N, Keyvani K, Pagenstecher A, Dammann P, Sandalcioglu IE, Sure U, et al. Implications of dll4-Notch signaling activation in primary glioblastoma multiforme. Neuro Oncol 2013;15:1366-78. [PUBMED]
19.	Li ZQ, Gong LL, Wen ZH, Wang J, Xu CS, Huang XD, et al. Delta-like ligand 4 correlates with endothelial proliferation and vessel maturation in human malignant glioma. Onkologie 2012;35:763-8.
20.	Qiu XX, Chen L, Wang CH, Lin ZX, Zhou CF, Liu SY, et al. High delta-like ligand 4 (DLL4) is correlated with peritumoral brain edema and predicts poor prognosis in primary glioblastoma. Medicine (Baltimore) 2014;93:e57. [PUBMED]
21.	Wen PY, Macdonald DR, Reardon DA, Cloughesy TF, Sorensen AG, Galanis E, et al. Updated response assessment criteria for high-grade gliomas: Response assessment in neuro-oncology working group. J Clin Oncol 2010;28:1963-72. [PUBMED]
22.	Li Z, Wang J, Gong L, Wen Z, Xu C, Huang X, et al. Correlation of delta-like ligand 4 (DLL4) with VEGF and HIF-1α expression in human glioma. Asian Pac J Cancer Prev 2011;12:215-8.
23.	Weidner N. Current pathologic methods for measuring intratumoral microvessel density within breast carcinoma and other solid tumors. Breast Cancer Res Treat 1995;36:169-80. [PUBMED]
24.	Hammoud MA, Sawaya R, Shi W, Thall PF, Leeds NE. Prognostic significance of preoperative MRI scans in glioblastoma multiforme. J Neurooncol 1996;27:65-73. [PUBMED]
25.	Sato H, Enmi J, Teramoto N, Hayashi T, Yamamoto A, Tsuji T, et al. Comparison of gd-DTPA-induced signal enhancements in rat brain C6 glioma among different pulse sequences in 3-tesla magnetic resonance imaging. Acta Radiol 2008;49:172-9. [PUBMED]
26.	Lobov IB, Renard RA, Papadopoulos N, Gale NW, Thurston G, Yancopoulos GD, et al. Delta-like ligand 4 (DLL4) is induced by VEGF as a negative regulator of angiogenic sprouting. Proc Natl Acad Sci U S A 2007;104:3219-24. [PUBMED]
27.	Li JL, Sainson RC, Shi W, Leek R, Harrington LS, Preusser M, et al. Delta-like 4 Notch ligand regulates tumor angiogenesis, improves tumor vascular function, and promotes tumor growth in vivo. Cancer Res 2007;67:11244-53. [PUBMED]
28.	Zhu TS, Costello MA, Talsma CE, Flack CG, Crowley JG, Hamm LL, et al. Endothelial cells create a stem cell niche in glioblastoma by providing NOTCH ligands that nurture self-renewal of cancer stem-like cells. Cancer Res 2011;71:6061-72. [PUBMED]
29.	Hurwitz H, Fehrenbacher L, Novotny W, Cartwright T, Hainsworth J, Heim W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004;350:2335-42. [PUBMED]
30.	Sandler A, Gray R, Perry MC, Brahmer J, Schiller JH, Dowlati A, et al. Paclitaxel-carboplatin alone or with bevacizumab for non-small-cell lung cancer. N Engl J Med 2006;355:2542-50. [PUBMED]