Inhibition of VEGF and Angiopoietin-2 to Reduce Brain Metastases of Breast Cancer Burden

Abstract

For metastases in the central nervous system, angiogenesis enhances metastatic potential and promotes progression. Primary factors which drive vessel growth are vascular endothelial growth factor (VEGF) and angiopoietin-2. Preclinical models show inhibition of either factor reduces metastases spread and inhibits growth. This work sets out to answer two questions in a preclinical mouse model. First, whether the combined inhibition of VEGF and angiopoietin-2, reduces passive permeability and limits drug uptake into brain metastases; and second, whether this inhibition reduces metastases burden in brain. We observed combinatorial inhibition of VEGF and angiopoietin-2, decreased (p < 0.05) angiogenesis and vascular branching in an aortic ring assay and decreased (p < 0.05) endothelial wound closure times. Using a brain metastases of breast cancer model (induced by intracardiac injections of brain seeking MDA-MB-231Br cells or 4T1Br cells), we observed, similar to VEGF, angiopoetin-2 expression correlates to increased angiogenesis (p < 0.05) and increased lesion permeability. To determine efficacy, animals were administered bevacizumab plus L1-10 (angiopoietin inhibitor) twice per week until neurological symptoms developed. Lesion permeability significantly decreased by ∼50% (p < 0.05) compared to untreated lesions, but remained ∼25% greater (p < 0.0%) than brain. In subsequent experiments, animals were administered similar regimens but sacrificed on day 32. The number of metastatic lesions developed was significantly (p < 0.001) reduced in the bevacizumab group (56%) and combination group (86%). Lesions' size was reduced in bevacizumab treated lesions (∼67%) and bevacizumab and L1-10 treated lesions (∼78%) developing area < 0.5 mm². In summary, combinatorial inhibition of VEGF and angiopoietin reduces lesion permeability and brain metastatic burden.

Keywords: L1-10; angiogenesis; bevacizumab; brain metastases; permeability; prevention.

FIGURE 1

Vascular endothelial growth factor (VEGF) and Ang-2 promote angiogenesis and can be modulated by inhibition. The endothelial branching observed from in vitro aortic ring angiogenesis assays in the presence of VEGF (10 ng/mL) + Ang-2 (100 ng/mL) and the addition of bevacizumab (20 μg/mL) and L1-10 (6.5 μg/mL) for 24, 48, and 72 h duration of treatment (A). An increase in endothelial branching compared to control is denoted by (^∗). A reduction in endothelial branching compared to VEGF + Ang-2 is denoted by (^†). Wound closure of bEND5 cells in the presence of Ang-2 (10 ng/mL) + VEGF (5 ng/mL) and the addition of bevacizumab (20 μg/mL) and L1-10 (6.5 μg/mL) in comparison to Ang-2 (10 ng/mL) and VEGF (5 ng/mL) over 24, 48, and 72 h (B). An increase in percent wound closure relative to control is denoted by (^∗) and a decrease in wound closure relative to Ang-2 + VEGF is denoted by (^†). Statistical evaluation was determined by student’s t-tests; ^∗ and ^† for p < 0.05, ^∗∗ and ^†† for p < 0.01, ^∗∗∗ and ^††† for p < 0.001 and ^∗∗∗∗ and ^†††† for p < 0.0001.

FIGURE 2

Expression of angiopoietin-2 is greater in lesions with higher permeability. Brain slices from un-treated metastases bearing animals were stained for the hypoxia induced vascular destabilizing protein Ang-2. Ang-2 expression in MDA-MB-231-BR-Her2 metastases is shown in non-permeable (A) and permeable (B) metastatic lesions. Representative images in the 4T1-BR5 model are shown in a lesion with low permeability (C) and a lesion with high permeability (D). Ang-2 (red), CD31 (green), DAPI (blue). Ang-2 intensity in permeable and non-permeable lesion in both preclinical models of brain metastases of breast cancer (E). Permeable lesions demonstrated greater Ang-2 expression than non-permeable in both models, with a >2-fold increase in the MDA-MB-231-BR-Her2 model, indicating the presence of vascular destabilization. Scale bar = 25 μm. Statistical significance was determined using Student’s t-test to determine difference in Ang-2 expression between the two permeability groups within each model (^∗∗∗ for p < 0.001).

FIGURE 3

Treatment of bevacizumab and L1-10AIB during metastasis development decreases permeability of ¹⁴C-AIB, which correlates with metastasis size. Representative brain slice image of enhanced green fluorescent protein (eGFP) labeled lesions (A) and corresponding quantitative autoradiography (QAR) image (B) with ¹⁴C-AIB distribution. The percent brain space of ¹⁴C-AIB in normal brain and metastases (C). Fold-increase in permeability with metastasis size reveals a weak correlation in bevacizumab plus L1-10 treated lesions (n = 65, r² = 0.162, p < 0.05) (D). Statistical significance was determined using ANOVA followed by Bonferroni’s multiple comparisons test for AIB brain space percentage, and linear regression analysis for fold increase in permeability versus lesion size (^∗∗∗ for p < 0.001).

FIGURE 4

Brain uptake of ¹⁴C-paclitaxel in brain metastases after bevacizumab treatment. Bevacizumab reduces (p < 0.05) metastasis paclitaxel uptake to a mean of 163 ng/g, ranging from 7.76 to 788 ng/g (n = 81) as compared to our previously reported value of 276 ng/g ranging from 2.56 to 1768 (Lockman et al., 2010). In addition, there was a strong correlation with Texas Red 3 kDa dextran permeability (r² = 0.72, p < 0.001) (A). In addition, paclitaxel concentrations in metastases from the bevacizumab group showed a weak correlation with lesion size (r² = 0.071, p < 0.05) (B).

FIGURE 5

Bevacizumab and L1-10 treatment reduces metastatic burden in brain. Representative images of MDA-MB-231-BR-Her2 non-treated metastatic brain (A), bevacizumab treated brain (B), and bevacizumab + L1-10 treated brain (C). Bevacizumab (10 mg/kg) or bevacizumab + L1-10 (4 mg/kg) was administered on day 10 after inoculation (with cancer cells and the number of metastatic lesions developed was significantly (p < 0.001) reduced in the bevacizumab alone group (56%) and in the combination group (86%) (D). Treatment also reduced lesion size with 67% ± 1 of bevacizumab treated lesions and 78% ± 10 of bevacizumab + L1-10 treated lesions developing area < 0.5 mm². Data illustrate a significant reduction in metastatic burden with simultaneous inhibition of the VEGF/angiopoietin-2 pathway. Significance was determined using ANOVA analysis followed by Bonferroni’s multiple comparisons test; ^∗ for p < 0.05, ^∗∗ for p < 0.01 and ^∗∗∗ for p < 0.001.)