Role of Vascular Endothelial Growth Factor in Ovarian Cancer

Materials

A mouse monoclonal antibody (A4.6.1) directed against human VEGF was used to neutralize VEGF activity in vivo. Characterization of this antibody, including its high specificity toward human VEGF and its ability to inhibit VEGF activity in vitro and in vivo, as well as to block binding of VEGF to its receptors in vivo, has been described previously. 46,62 The antibody does not inhibit the activity of mouse VEGF (unpublished data). The human ovarian cystadenocarcinoma cell line, SKOV-3, was obtained from the American Type Culture Collection (Manassas, VA). One-month-old female immunodeficient mice (BALB/c nu/nu) were obtained from Simonsen Laboratories (Gilroy, CA), housed in isolated conditions, and fed autoclaved standard pellets and water. All protocols involving immunodeficient mice were approved by the Committee on Animal Care, University of California, San Francisco.

Cell Culture

The SKOV-3 cells were cultured in Dulbecco’s modified Eagle’s medium H-21 containing 10% fetal calf serum, glucose (4.5 g/L), penicillin G (100 U/ml), streptomycin (2.5 μg/ml), glutamine (2 mmol/L), and fungizone (2.5 μg/ml). All cell culture reagents were obtained from the Cell Culture Facility, University of California, San Francisco. Before in vivo inoculation, SKOV-3 cells were grown to confluence, harvested by trypsinization, and resuspended in Ca²⁺/Mg²⁺-free phosphate buffered saline (PBS). In preliminary studies, we determined that SKOV-3 cells express VEGF in vivo and in vitro using reverse transcription-polymerase chain reaction and immunocytochemistry, respectively, and that A4.6.1 does not affect their proliferation in vitro (data not shown).

In Vivo Inoculation of SKOV3 Cells

The SKOV-3 cells were prepared for inoculation as described above and injected as a bolus either into the peritoneum (IP group; n = 31; 10 × 10 6 cells per mouse in 200 μl of PBS) or into the dorsal subcutaneous tissue (SC group; n = 8; 5 to 10 × 10 6 cells in 50 μl of PBS) of athymic mice. Some SC mice received two boluses of SKOV-3 cells, one in each flank. Seven to 10 days after SKOV-3 inoculation, some of the mice (IP group, n = 16; SC group, n = 5) were treated with A4.6.1 (100 μg in 0.1 ml of PBS, intraperitoneally, twice per week), and the rest were treated with the same volume of vehicle. A4.6.1 treatment was delayed to ensure that tumor engraftment was not inhibited. The size of subcutaneous tumors was measured twice weekly using calipers fitted with a Vernier scale. For each tumor, two perpendicular measurements were obtained from which an estimate of tumor radius was derived. Tumor volume was then calculated based on the assumption that tumors were spherical. At the end of the experimental period, mice were killed by anesthetic overdose. At autopsy, SKOV-3 tumors were excised, fixed in 4% paraformaldehyde/100 mmol/L PBS, pH 7.4, at 4°C for 24 hours, and embedded in paraffin. Paraffin sections (10 μm) were used for histochemical analysis.

Statistics

Data were analyzed using the unpaired Student’s t-test for statistical comparison between groups. Differences between groups were considered statistically significant at P < 0.05. Experiments were performed in triplicate.

Experimental Model of Human Ovarian Cancer

All IP mice receiving PBS treatment (ie, control mice) developed a swollen abdomen, indicative of ascites formation and intraperitoneal carcinomatosis, within 3 to 6 weeks of SKOV-3 administration. Soon after (within 5 to 7 days) the appearance of abdominal swelling, PBS-treated IP mice became cachectic and as a consequence were euthanized in accordance with the animal care protocol. The intraperitoneal carcinomatosis in the IP mice closely resembled peritoneal metastases from poorly differentiated stage III papillary serous ovarian cancer. 63 The neoplasms were characterized by trabecular and solid patterns of growth with variable degrees of cytological atypia. Numerous mitotic figures were identified, and papillary growth was common on the luminal side. In most tumors, desmoplastic stroma typical of ovarian cancer metastases was seen. Tumors were found on the surfaces of the peritoneum, diaphragm, intestines (Figure 1) ▶ , uterus and associated fat, and stomach. Tumors were rarely found on the liver or spleen, and there was no evidence of visible metastasis to organs outside of the peritoneum. In general, the tumors did not invade the host tissue to which they were adherent; however, some focal invasion into muscle was seen. Extensive tumor growth was detected in the uterine fat, and foci of tumor were observed in the uterine lymphatics. The sites and extent of the peritoneal carcinomatosis and the production of ascitic fluid induced by intraperitoneal administration of SKOV-3 cells in immunodeficient mice were similar to those seen in women with ovarian epithelial cancer.

Role of VEGF in SKOV-3 Tumor Growth

To examine directly the role of tumor-derived VEGF in SKOV-3 tumor growth, we established an in vivo model in which SKOV-3 tumors were grown subcutaneously in immunodeficient mice. Well defined subcutaneous tumors developed within 7 days of SKOV-3 inoculation and were of sufficient size to permit accurate measurement. Tumor growth was rapid, and within 3 weeks the subcutaneous foci were 5 to 10 mm in diameter and began to exhibit vascular islands that eventually formed blood-filled cysts. By 6 weeks, the largest tumors were approximately 20 mm in diameter and contained numerous cysts that eventually ruptured. At this stage of tumor progression, the mice were killed.

To assess the role of VEGF in tumor growth, some SC mice (n = 5 mice; 11 tumors) were treated with A4.6.1. Control mice (n = 3 mice; 6 tumors) were treated with PBS. In preliminary studies, we found that treatment with a nonspecific antibody of the same IgG type had no effect on tumor growth and was essentially equivalent to vehicle alone (data not shown). A4.6.1 significantly inhibited the growth of subcutaneous SKOV-3 tumors within 2 weeks of treatment (Figure 2) ▶ ; tumor size did not progress beyond the size attained at the initiation of A4.6.1 treatment and the tumors did not form cysts. After cessation of A4.6.1 treatment (Figure 2B) ▶ , tumor growth resumed, and within 2 weeks the tumors developed blood-filled cysts and the mice had to be killed. In preliminary studies in which fewer cells were used for subcutaneous inoculation, after discontinuance of A4.6.1 treatment, tumor growth resumed and within 3 weeks the rate of growth paralleled that of controls. Tumor growth also was inhibited when A4.6.1 was administered late in tumor progression (30 days) (Figure 2B) ▶ . Interestingly, in these tumors A4.6.1 appeared to deplete the contents of the cysts; their volume was markedly reduced, and some cysts involuted without rupture. In addition, tumors lost their red coloration and became more skin toned in appearance.

There was no evidence that A4.6.1 induced an antibody-dependent cellular cytotoxicity or a macrophage-mediated response that could have inhibited tumor growth. Histological examination of growth-inhibited subcutaneous tumors failed to demonstrate any significant inflammatory response (data not shown).

Role of VEGF in Intraperitoneal Tumor Growth

Studies of subcutaneous SKOV-3 tumor growth established the pivotal role of VEGF in tumor progression. However, as ovarian cancer is not a subcutaneous malignancy, we studied the effects of A4.6.1 treatment on the growth and progression of intraperitoneal SKOV-3 tumors. IP mice were treated with A4.6.1 (n = 16) or PBS (n = 15) in an identical fashion to those bearing subcutaneous tumors. However, intraperitoneal tumor growth could not be monitored directly and, because of its spread within the abdomen, could not be quantified accurately. Therefore, intraperitoneal tumor burden was assessed qualitatively at postmortem examination. In all animals, treatment was initiated 8 days after SKOV-3 inoculation and continued for various times.

In experiment 1 (Figure 3A) ▶ , all animals were killed 21 days after SKOV-3 inoculation. At postmortem examination, two of the three PBS-treated animals exhibited abdominal swelling with a moderate level of ascites and had moderate and easily detectable intraperitoneal tumor burden. The remaining PBS-treated animal had no detectable abdominal swelling or ascites and only a mild tumor burden. None of the A4.6.1-treated IP animals showed signs of ascites formation or cachexia at the time of postmortem examination, and intraperitoneal SKOV-3 tumor burden was barely detectable. Interestingly, A4.6.1 inhibited the growth of small subcutaneous tumors at the site of SKOV-3 injection that developed in some IP animals.

In experiment 2 (Figure 3B) ▶ , the effects of prolonged A4.6.1 treatment (n = 6) were examined. All of the PBS-treated animals (n = 6) displayed signs of ascites formation and cachexia at various times after SKOV-3 inoculation; the first PBS-treated animal was killed at 4.5 weeks and the last at 6.5 weeks because of severe cachexia associated with the developing ascites. One A4.6.1-treated animal exhibited ascites and cachexia and therefore was killed at 5.5 weeks. This mouse had a high tumor burden. Four of the A4.6.1-treated animals were killed at 6.5 weeks. The remaining A4.6.1-treated IP mouse was killed the following day. Five of the six A4.6.1-treated IP animals appeared normal at the time of postmortem examination, and no ascites formation could be detected. In A4.6.1-treated mice, intraperitoneal tumor burden was variable: in three of the six animals it was high and comparable with tumor burden in those mice receiving PBS alone, whereas in the other IP animals tumor burden ranged from moderate to scarce. The reason for the apparent lack of response in the one A4.6.1-treated mouse is uncertain.

In experiment 3 (Figure 3C) ▶ , A4.6.1 treatment was stopped after the last PBS-treated animal was killed (approximately 7 weeks after SKOV-3 inoculation). At that time, three of the six A4.6.1-treated animals were killed. As with experiment 2, all of the PBS-treated animals exhibited abdominal swelling and became cachectic approximately 6 weeks after receiving SKOV-3 cells. All of these animals had severe ascites and a high intraperitoneal tumor burden. In contrast, the three A4.6.1-treated animals killed at 6.5 weeks had no detectable ascites and a variable, but clearly detectable, tumor burden. Importantly, within 2 to 3 weeks of cessation of A4.6.1 treatment, the remaining mice developed severe ascites, became cachectic, and had to be killed. The tumor burden in these animals varied from moderate to high.