Vascular endothelial growth factor is a promising therapeutic target for the treatment of clear cell carcinoma of the ovary

Abstract

This study examines the role of vascular endothelial growth factor (VEGF) as a therapeutic target in clear cell carcinoma (CCC) of the ovary, which has been regarded as a chemoresistant histologic subtype. Immunohistochemical analysis using tissue microarrays of 98 primary ovarian cancers revealed that VEGF was strongly expressed both in early-stage and advanced-stage CCC of the ovary. In early-stage CCCs, patients who had tumors with high levels of VEGF had significantly shorter survival than those with low levels of VEGF. In vitro experiments revealed that VEGF expression was significantly higher in cisplatin-refractory human CCC cells (RMG1-CR and KOC7C-CR), compared with the respective parental cells (RMG1 and KOC7C) in the presence of cisplatin. In vivo treatment with bevacizumab markedly inhibited the growth of both parental CCC cell-derived (RMG1 and KOC7C) and cisplatin-refractory CCC cell-derived (RMG1-CR and KOC7C-CR) tumors as a result of inhibition of tumor angiogenesis. The results of the current study indicate that VEGF is frequently expressed and can be a promising therapeutic target in the management of CCC. Bevacizumab may be efficacious not only as a first-line treatment but also as a second-line treatment of recurrent disease in patients previously treated with cisplatin.

Figure 1. VEGF is frequently expressed in both ovarian clear cell carcinomas (CCCs) and serous denocarcinomas (SACs)

A, Representative photographs of clear cell carcinoma (CCC), and serous adenocarcinoma (SAC). (hematoxylin-eosin staining, × 400). (i), Clear cell carcinoma charaterized by clear cells containing abundant cytoplasmic glycogen. (ii), Serous adenocarcinoma characterized by an exophytic proliferation of cellular papillae. B, Representative photomicrographs of weak, moderate, and strong staining for VEGF are shown. Magnifications: ×100, and ×400 (inset). C, Kaplan-Meier curves comparing the overall survival of the patients in all stages. The overall survival in the weak expression-group was significantly longer than that in the strong-expression group (log-rank; p=0.0124). The overall survival in patients with advanced stage CCCs was significantly shorter than that observed in the early stage, weak expression-group (log-rank; p<0.0001) or in the early stage, strong-expression group (log-rank; p=0.0036).

Figure 1. VEGF is frequently expressed in both ovarian clear cell carcinomas (CCCs) and serous denocarcinomas (SACs)

A, Representative photographs of clear cell carcinoma (CCC), and serous adenocarcinoma (SAC). (hematoxylin-eosin staining, × 400). (i), Clear cell carcinoma charaterized by clear cells containing abundant cytoplasmic glycogen. (ii), Serous adenocarcinoma characterized by an exophytic proliferation of cellular papillae. B, Representative photomicrographs of weak, moderate, and strong staining for VEGF are shown. Magnifications: ×100, and ×400 (inset). C, Kaplan-Meier curves comparing the overall survival of the patients in all stages. The overall survival in the weak expression-group was significantly longer than that in the strong-expression group (log-rank; p=0.0124). The overall survival in patients with advanced stage CCCs was significantly shorter than that observed in the early stage, weak expression-group (log-rank; p<0.0001) or in the early stage, strong-expression group (log-rank; p=0.0036).

Figure 2. VEGF production in CCC cells and anti-angiogenic activity of bevacizumab In vitro

A, 5×10⁴ RMG1 and KOC7C cells were incubated in DMEM Ham's F-12 containing 1% FBS for 24 h. Levels of secreted VEGF protein in the conditioned media under normoxic (20% O₂) or hypoxic (1% O₂) conditions were measured by ELISA assay. *, p< 0.05, significantly different from control. B, The effect of either recombinant VEGF or cultured medium on the proliferation of HUVECs. HUVECs were treated with VEGF or cultured medium containing VEGF secreted by CCC cells (RMG1 or KOC7C, under hypoxic conditions) with or without the indicated concentration of bevacizumab in the presence of 5% fetal bovine serum (FBS) for 72h under normoxic (20% O₂) condition. Cell viability was assessed by MTS assay. *, p< 0.05, significantly different from control. C. Inhibitory effect of bevacizumab on VEGF-induced tube formation in vitro. HUVECs were cultured with VEGF or cultured with medium containing VEGF secreted by CCC cells (RMG1 or KOC7C, under hypoxic conditions), with or without bevacizumab for 8 h under normoxic (20% O₂) conditions. (l), Representative photographs of tube formation on matrigel are shown. a, PBS; b, Bevacizumab; c, VEGF; d, VEGF plus Bevacizumab. (II, III), Three random fields per sample were recorded and tube number of every field was measured. Columns, mean (n=4); bars, SD. The concentrations of agents used in this experiment are as follows: VEGF; 1 ng/ml, bevacizumab; 10 µg/ml. *, p< 0.05, significantly different from control.

Figure 2. VEGF production in CCC cells and anti-angiogenic activity of bevacizumab In vitro

A, 5×10⁴ RMG1 and KOC7C cells were incubated in DMEM Ham's F-12 containing 1% FBS for 24 h. Levels of secreted VEGF protein in the conditioned media under normoxic (20% O₂) or hypoxic (1% O₂) conditions were measured by ELISA assay. *, p< 0.05, significantly different from control. B, The effect of either recombinant VEGF or cultured medium on the proliferation of HUVECs. HUVECs were treated with VEGF or cultured medium containing VEGF secreted by CCC cells (RMG1 or KOC7C, under hypoxic conditions) with or without the indicated concentration of bevacizumab in the presence of 5% fetal bovine serum (FBS) for 72h under normoxic (20% O₂) condition. Cell viability was assessed by MTS assay. *, p< 0.05, significantly different from control. C. Inhibitory effect of bevacizumab on VEGF-induced tube formation in vitro. HUVECs were cultured with VEGF or cultured with medium containing VEGF secreted by CCC cells (RMG1 or KOC7C, under hypoxic conditions), with or without bevacizumab for 8 h under normoxic (20% O₂) conditions. (l), Representative photographs of tube formation on matrigel are shown. a, PBS; b, Bevacizumab; c, VEGF; d, VEGF plus Bevacizumab. (II, III), Three random fields per sample were recorded and tube number of every field was measured. Columns, mean (n=4); bars, SD. The concentrations of agents used in this experiment are as follows: VEGF; 1 ng/ml, bevacizumab; 10 µg/ml. *, p< 0.05, significantly different from control.

Figure 3. Effect of bevacizumab on the growth of ovarian CCC in vivo

Athymic nude mice were inoculated s.c. with RMG1 or KOC7C cells, with 12 mice in each group. When tumors reached ~50 mm³, the mice were treated with placebo or 5 mg/kg bevacizumab twice a week for 4 wks. A and B, graphs depicting weekly tumor volumes (mm³) for each treatment group. Points, mean; bars, SD. *, p< 0.05, significantly different from PBS-treated mice. C, To determine the anti-angiogenic activity of bevacizumab, the microvessel area (MVA) in the RMG1-derived subcutaneous tumors was assessed by immunohistochemistry with anti-CD31/PECAM-1 antibody. D, The area of microvessels in the RMG1-derived subcutaneous tumors was measured, and percent area was calculated in each group. Columns, mean; bars, SD. *, p< 0.05, significantly different from PBS-treated tumors.

Figure 4. VEGF expression and its role as a therapeutic target in cisplatin-refractory CCC cells

(A, B), Establishment of cisplatin-refractory variant cell lines. Cisplatin-refractory sublines were established as described in “Materials and Methods”. (A), Parental (KOC7C and RMG1) and cisplatin-refractory variant (KOC7C-CR and RMG1-CR) cells were treated with the indicated concentrations of cisplatin in the presence of 5% FBS for 72 h. Cell viability was assessed by MTS assay. Points, mean; bars, SD (*, p< 0.05, **, p<0.01, significantly different from control.). (B), Effect of cisplatin on cleavage of PARP in parental and cisplatin-refractory variant cell lines. KOC7C, KOC7C-CR, RMG1 and RMG1-CR treated with 10 µM cisplatin or bevacizumab for 24 h. Cells were harvested, and then lysates were subjected to Western blotting using anti-PARP or anti-β-actin antibody. (C), Production of VEGF in cisplatin-refractory sublines and parental chemosensitive cells. Levels of secreted VEGF protein in the conditioned media under normoxic or hypoxic condition were measured by ELISA assay. Columns, mean; bars, SD. *, p< 0.05, significantly different from control.

Figure 4. VEGF expression and its role as a therapeutic target in cisplatin-refractory CCC cells

(A, B), Establishment of cisplatin-refractory variant cell lines. Cisplatin-refractory sublines were established as described in “Materials and Methods”. (A), Parental (KOC7C and RMG1) and cisplatin-refractory variant (KOC7C-CR and RMG1-CR) cells were treated with the indicated concentrations of cisplatin in the presence of 5% FBS for 72 h. Cell viability was assessed by MTS assay. Points, mean; bars, SD (*, p< 0.05, **, p<0.01, significantly different from control.). (B), Effect of cisplatin on cleavage of PARP in parental and cisplatin-refractory variant cell lines. KOC7C, KOC7C-CR, RMG1 and RMG1-CR treated with 10 µM cisplatin or bevacizumab for 24 h. Cells were harvested, and then lysates were subjected to Western blotting using anti-PARP or anti-β-actin antibody. (C), Production of VEGF in cisplatin-refractory sublines and parental chemosensitive cells. Levels of secreted VEGF protein in the conditioned media under normoxic or hypoxic condition were measured by ELISA assay. Columns, mean; bars, SD. *, p< 0.05, significantly different from control.

Figure 5. Effect of bevacizumab on the growth of cisplatin-refractory CCC-derived tumors in vivo

Athymic nude mice inoculated s.c. with KOC7C-CR cells (n=12) or RMG1-CR cells (n=12) were treated with placebo or 5 mg/kg bevacizumab twice a week for 4 wks. A and B, graphs depicting weekly tumor volumes (mm³) for each treatment group. Points, mean; bars, SD. *, p< 0.05, significantly different from PBS-treated mice.

Figure 6. Comparison between the effect of bevacizumab versus cisplatin on the growth of CCC-derived tumors in vivo

Athymic nude mice were inoculated s.c. with RMG1 cells or RMG1-CR cells, with 12 mice in each group. When tumors reached ~50 mm³, mice were treated with PBS, 5 mg/kg bevacizumab, or 3 mg/kg cisplatin twice a week for 4 wks. PBS, bevacizumab, and cisplatin were administered intraperitoneally. A and B, graphs depicting weekly tumor volumes (mm³) for each treatment group. Points, mean; bars, SD. *, p< 0.05, significantly different from PBS-treated mice.