Abstract
Background: Potential tissue and serum biomarkers were assessed for predicting efficacy of bevacizumab in ovarian cancer (OC). Patients and Methods: Twenty patients with OC received chemotherapy alone (control group) or in combination with bevacizumab (study group) in this non-randomised study. Pre- and post-treatment serum concentrations of vascular endothelial growth factor (VEGF), 8-hydroxydeoxyguanosine (8-OHdG) and tissue inhibitor of metalloproteinase-1 (TIMP-1) were measured in all patients. In addition, immunohistochemical expressions VEGF receptors (R1, R2), hypoxia-inducible factor 1α (HIF-1 α), matrix metalloproteinases (-2, -9) and TIMP-2 were analysed in tumours from bevacizumab-treated patients. Results: 8-OHdG and HIF-1α immunostainings were more highly expressed among patients with progression-free survival (PFS) >23 months than in patients with PFS <12 months. Similarly, MMP-9 was more frequently highly expressed in those with long versus short PFS. Serum VEGF concentrations decreased substantially in all bevacizumab-treated patients versus only 20% of the control group. Conclusion: Our results indicate differences in MMP-9 and HIF-1α expression in relation to duration of PFS and effects on serum VEGF when bevacizumab is used in combination with chemotherapy.
- Targeted therapy
- VEGF
- matrix metalloproteinase
- oxidative stress
- reactive oxygen species
- angiogenesis
- bevacizumab
Vascular endothelial growth factor (VEGF) is one of the most potent angiogenesis-inducing factors and appears to be an indicator of angiogenic potential in tumours. Bevacizumab (a monoclonal antibody for VEGF) has demonstrated a significant survival benefit in cancer of colon, renal, kidney and lung and significantly increased progression-free survival (PFS) in metastatic breast cancer (1). In ovarian cancer, bevacizumab either alone (2-4) or in combination with various cytotoxic compounds (5-9) has shown promising activity in relapsed and platinum-resistant ovarian cancer. Bevacizumab has shown efficacy as first-line treatment for ovarian cancer (10) and is being evaluated in two large, randomised phase III trials in the first-line setting (GOG 218 and ICON7).
Another important avenue of research relates to the transcription factor hypoxia-inducible factor 1 (HIF-1). Hypoxia is an important regulator of angiogenesis and several stress-response genes are activated under hypoxic conditions, especially in malignant cells. HIF-1α regulates angiogenesis and VEGF expression (11). Reactive oxygen species (ROS) formation and subsequent oxidative stress are closely linked to hypoxic circumstances. ROS participate in downstream signalling pathways of HIF-1 and play an essential role as hypoxia sensors of HIF-1 (12). 8-Hydroxydeoxyguanosine (8-OHdG) indicates total ROS production and is a widely used marker of ROS-induced DNA damage, which can be assessed with both immunohistochemical (IHC) and enzyme-linked immunosorbent assay (ELISA) techniques. In our recently reported study, high 8-OHdG tumour expression correlated with worse prognosis in patients with epithelial ovarian cancer (13).
Serum VEGF concentrations correlate with overall survival in ovarian cancer (14). Experiments in ovarian cancer cell cultures suggest that the invasion potential of cancer cells induced by VEGF is mediated by matrix metalloproteinases (MMPs). MMP-2 and MMP-9 play an important role in tumour progression and metastasis in numerous malignancies (15-16) and are implicated in both pro- and anti-angiogenic processes. Several in vitro and in vivo studies have shown a relationship between regulation of VEGF by MMPs and their tissue inhibitors (TIMP-1 and TIMP-2) (15-17). VEGF inhibition down-regulates the expression of MMP-9 (18) and conversely, MMP-9 mediates the release of VEGF from matrix-associated extracellular sources (19). TIMP-2 inhibits the response of endothelial cells to VEGF stimulation (20).
Details of antigens and antibodies used.
In the present study, IHC expression of VEGF, its receptors VEGFR1 and VEGFR2, oxidative stress marker 8-OHdG, HIF-1α and molecules involved in metastasis and invasion processes (MMP-2, MMP-9, TIMP-1 and TIMP-2) were evaluated in patients receiving bevacizumab in combination with chemotherapy for ovarian cancer, with the aim of identifying ovarian cancer patients deriving the greatest benefit from bevacizumab therapy. In addition, serum concentrations of VEGF, 8-OHdG and TIMP-1 were evaluated before and after bevacizumab plus chemotherapy (study group) and compared with concentrations in patients treated with chemotherapy alone.
Patients and Methods
Patient population. A total of 21 patients with stage III serous ovarian cancer participating in a prospective, open, single-arm phase II trial (FINAVAST, trial number ML18583) were eligible to participate in the present biomarker study. After surgery, patients were treated with 6-9 cycles of carboplatin area under curve (AUC) 5, paclitaxel 175 mg/m2 and bevacizumab 7.5 mg/kg every three weeks, after which bevacizumab treatment was continued until progression, or for up to 12 cycles. The bevacizumab treatment was started from the second chemotherapy cycle on. Response was evaluated by computed tomography (CT) using RECIST (21) or by cancer antigen (CA) 125 using the Rustin criteria (22). When disease progression was suspected by CA125 increase, progression was to be confirmed by CT scan. PFS was defined as a time from surgery to disease progression. Patients gave written consent to provide tumour samples and serum samples, which were taken before and after 6-9 cycles of chemotherapy. The study was approved by the Ethical Committee of Pirkanmaa Hospital District, Tampere, Finland.
The control group used for comparative serum analysis consisted of ten patients with stage III-IV serous ovarian cancer who were treated with paclitaxel and carboplatin according to the standard ovarian cancer treatment programme in Oulu University Hospital.
Immunohistochemistry. Ovarian cancer tissue samples were fixed in neutral buffered formalin and embedded in paraffin. Sections (5 μm) were cut from the specimens and placed on glass slides (Menzel-Gläser, Germany). The sections were first de-paraffinised in xylene and re-hydrated in a descending ethanol series, incubated in 10 mM citrate buffer (pH 6.0), boiled in a microwave oven for 10 minutes, and cooled at room temperature before adding the primary antibody (Table I). Negative controls were prepared using the same procedure except that the primary antibodies were replaced by phosphate-buffered saline (PBS) and serum isotype controls (Zymed Laboratories, Inc., South San Francisco, CA, USA). Immunoreactivity in the samples was assessed semi-quantitatively by grading both the staining intensity in the tumour cells and the proportion of positively-stained tumour cells. Immunoreactivity was evaluated by assigning the staining reaction to one of three groups: 0=no staining intensity; 1=weak or moderate staining intensity (20-50% of cells); 2=strong staining intensity (>50% of cells) (Figure 1). The IHC staining was independently assessed by two authors (P.K., U.P.).
Serum measurements of TIMP-1, VEGF, 8-OHdG. Serum TIMP-1 concentrations were analysed using microtiter plates (EIA/RIA plate, 96-well; Corning Incorporated, Corning, NY, USA) that were coated with monoclonal antibodies specific for TIMP-1 (MAB970; R&D Systems, MN, USA). The results were evaluated by Windows-based Control and Evaluation Software for Rosys Anthos microplate readers (Anthos Labtec Instruments, Wals, Austria). Serum VEGF concentrations were analysed with Quantikine VEGF Immunoassay (cat. no. DVE00; R&D Systems), which is a 4.5-hour solid-phase ELISA designed to measure VEGF165 concentrations in cell culture supernatants, serum and plasma. The manufacturer claims inter-assay and intra-assay variabilities of 6.2-8.8% and 5.1-6.7%, respectively. For determination of 8-OHdG concentrations, the Highly Sensitive 8-OHdG Check ELISA kit from the Japan Institute for the Control of Aging, (Fukuroi, Japan) was used. Assays were performed according to the manufacturers' instructions with few modifications. Duplicates of each sample were assayed and the outermost wells of the microtiter plate were also used. Forty samples were assayed in one plate. If concentrations from duplicate samples differed by more than 10%, the assay was repeated.
Strong 8-OHdG immunostaining can be seen in the nuclei of the ovarian carcinoma cells (A) (magnification ×105). HIF-1α shows only nuclear staining (B) (magnification ×210). Moderate MMP-9-positivity is found in cytoplasm of tumour cells (C) (magnification ×105).
Distribution of HIF1-α and 8-OHdG immunohistochemical staining in patients treated with paclitaxel, carboplatin and bevacizumab; patients with PFS <12 months versus >23 months.
Statistical analyses were conducted using SPSS 16.0 for Windows (Chicago, IL, USA). The significance of the associations was defined with Mann-Whitney U-test. Probability p-values <0.05 were considered significant.
Results
Clinical data. Baseline characteristics are shown in Table II. All patients had serous epithelial ovarian cancer. In the study group (patients treated with bevacizumab plus chemotherapy), the median PFS was 20.2 months (range 7-39 months). Patients were divided into two categories: those with PFS >23 months and those with PFS <12 months. We used these cut-offs because they clearly divided patients into two separate groups, responders and non-responders. In the control group treated with chemotherapy alone, the median PFS was 12.1 months. Only one patient had PFS >23 months.
Distribution of MMP-9 immunohistochemical staining in patients treated with paclitaxel, carboplatin and bevacizumab; patients with PFS <12 months versus >23 months.
Immunohistochemical data. IHC staining for MMP-2, MMP-9, TIMP-1, TIMP-2, VEGFR-1, VEGFR-2, 8-OHdG and HIF-1α was performed only in the 10 patients treated with bevacizumab combined with chemotherapy (study group), with comparison between those with long (>23 months) PFS and short PFS (<12 months). HIF-1α and 8-OHdG were more frequently overexpressed in those with sustained versus shorter response (Figure 2). Similarly, high (2++) MMP-9 staining was more frequent in patients with long versus short PFS (Figure 3). In contrast, high immunostaining was less frequent among patients with long versus short PFS for TIMP-1 and TIMP-2 staining profile (Table III).
Change in serum VEGF concentrations during the treatment of patients in the study group (taxane, platinum and bevacizumab-treated patients, light grey) and patients in the control group (taxane and platinum-treated patients, dark grey).
Patient characteristics of the study and control groups.
Immunohistochemical staining results in the study group (bevacizumab-treated patients): number of highly positive (++) staining in patients with PFS >23 months versus those with PFS <12 months.
Change in serum 8-OHdG concentrations during the treatment of patients in the study group (taxane, platinum and bevacizumab-treated patients, light grey) and patients in the control group (taxane and platinum-treated patients, dark grey).
Serum measurements. Both pre- and post-treatment blood samples were available for 7 out of the 10 patients treated with bevacizumab in combination with chemotherapy and all 10 patients receiving chemotherapy alone. There were marked decreases in serum VEGF concentrations in all patients receiving bevacizumab-containing therapy. On average, post-treatment values were 26% (range 5-79%) of the pretreatment values (Figure 4). There was no difference in the magnitude of the decrease according to duration of PFS. Among the 10 patients receiving chemotherapy alone, the post-treatment concentrations of VEGF decreased in only 2 out of 10 patients and increased in 8 out of 10 patients (Figure 4). Furthermore, in the two patients with a decrease in VEGF, the change was significantly smaller than in the bevacizumab-treated group (p=0.003). Serum 8-OHdG concentrations increased in six out of seven patients in the study group after bevacizumab treatment but there was no apparent difference between patients with long versus short PFS (Figure 5). In the control group, serum 8-OHdG concentrations increased in five out of nine patients (Figure 5). Serum TIMP-1 concentrations decreased in four out of seven patients in the study group, with no difference according to duration of PFS, and in three out of ten patients in the chemotherapy alone group (Figure 6).
Discussion
Serum VEGF concentrations were similar in the study population and control group in our study, suggesting that this population of patients with serous histology and advanced stage (III-IV) was relatively homogeneous. The role of bevacizumab in the treatment schedule was clearly indicated by the notable decrease in VEGF serum concentrations, contrasting significantly with the serum concentrations in patients receiving chemotherapy alone. However, there was no clear relationship between the magnitude of the decrease in VEGF serum concentrations and duration of PFS in the bevacizumab-treated group. In our study, the lower MMP-9 expression in patients with short PFS may reflect an alternative angiogenesis route independent of bevacizumab-induced VEGF inhibition, such as signalling by MMP-9.
Change in serum TIMP-1 concentrations during the treatment of patients in the study group (taxane, platinum and bevacizumab-treated patients, light grey) and in the control group (taxane and platinum-treated patients, dark grey).
IHC expressions of both hypoxia marker HIF-1α and oxidative stress marker 8-OHdG were higher in patients achieving a more prolonged response (median PFS >23 months) with bevacizumab therapy. This trend was particularly pronounced for HIF-1α: patients with PFS <12 months) showed no HIF-1α expression, whereas among patients with PFS >23 months, three out of five patients had evidence of HIF-1α expression. Consistent with this observation, IHC 8-OHdG expression was stronger in patients with longer PFS, and intensive 8-OHdG expression (++) appeared to predict good response to bevacizumab treatment. Pretreatment serum 8-OHdG concentrations did not appear to differentiate patients with long PFS from those with short PFS in the bevacizumab group. However, serum 8-OHdG concentrations increased after bevacizumab therapy in six out of the seven bevacizumab-treated patients (compared with five out of nine patients in the control group), which may suggest increased oxidative DNA damage associated with bevacizumab-containing therapy. These results are therefore consistent with previous preclinical studies, which have demonstrated ROS- and hypoxia-mediated VEGF overexpression (23), and might suggest that IHC assessment of 8-OHdG and HIF-1α may be useful when determining the potential benefit of bevacizumab treatment.
Bevacizumab is a promising anticancer agent for gynaecological malignancies and is under evaluation in numerous ongoing trials in ovarian cancer, both in advanced disease and in the first-line setting. Although the present study is limited by its small sample size, it nevertheless indicates differences in MMP-9 and HIF-1α expression in relation to duration of PFS and effects on serum VEGF when bevacizumab is used in combination with chemotherapy. The clinical benefit of the above markers for selecting bevacizumab treatment requires further evaluation in larger studies.
Acknowledgements
We would like to thank Ms Anne Bisi for her technical assistance and Hoffmann-La Roche Ltd for their donation of bevacizumab. This study was supported by grants from The Cancer Society of Northern Finland.
Footnotes
- Received December 6, 2009.
- Revision received February 22, 2010.
- Accepted February 23, 2010.
- Copyright© 2010 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
