The predictive value of serum VEGF in multiresistant ovarian cancer patients treated with bevacizumab.

ABSTRACT

OBJECTIVE: Bevacizumab, a humanized monoclonal antibody against VEGF (vascular endothelial growth factor), has shown antitumor activity, but so far no biomarkers have been identified to predict outcome. The purpose of the present study was to investigate the efficacy of bevacizumab in patients with multiresistant ovarian cancer and, furthermore, to investigate the possible predictive value of serum VEGF, VEGFR1-2 and VEGF gene polymorphisms. METHODS: Patients received single-agent bevacizumab 10 mg/kg every 3 weeks. All patients were followed with CA 125 measurements and serum VEGF/VEGFR1-2 levels prior to each cycle. Endpoints were response rate (RR), progression-free survival (PFS) and overall survival (OS). RESULTS: Thirty-eight patients were included. All patients were heavily pre-treated with a median of five prior regimens. The median number of bevacizumab treatments was 4. Overall response rate was 30% according to CA 125 (GCIG criteria). Median PFS was 5.9 months (95% CI, 3.5-9.4) and median OS was 8.6 months (95% CI, 6.6-12.8). The VEGF serum level decreased during treatment in all patients. A low pre-treatment VEGF level was predictive to response. The median value was 540 pg/ml and divided the patients into two groups with a response rate of 60% and 0%, respectively (p=0.0007). The difference translated to a significant difference in PFS (p=0.047) and OS (p=0.01). VEGF gene polymorphisms -2578, -1154, -460, +405, +936 did not reveal any association with response or survival and the same applied to serum VEGFR1-2. CONCLUSIONS: Single agent bevacizumab has activity in ovarian cancer patients. Pre-treatment serum VEGF seems to have predictive value.