Molecular correlates (EGFR status) and plasma VEGF levels associated with a phase II study of bevacizumab plus erlotinib (BE) for patients with recurrent ovarian cancer (OC) and fallopian tube (FT) cancer

Abstract

5554 Background: The epidermal growth factor receptor (EGFR) and vascular endothelial growth factor (VEGF) are commonly over- expressed in OC and may correlate with poor prognosis. EGFR mutations, although rare, have recently been reported in ovarian cancer. We examined these factors for prognostic value in the setting of treatment with BE. Results of a phase II trial of BE in recurrent OC pts have previously been reported; there were 1 CR and 1 PR among 13 patients. The trial was closed to accrual due to 2 bowel perforations. Methods: Pretreatment plasma VEGF, urine VEGF, and serum VEGFR2 levels are available on 9, 8, and 8 pts to date. VEGF levels were determined using an ELISA (R& D Systems, Minneapolis, MN). Tumor from 8 pts: 1(CR), 1 (PR), 5 (SD), 1 (PD) was available for immunohistochemical analysis for EGFR. Genomic DNA was successfully isolated from 6 paraffin embedded tumor specimens. EGFR exons 18 to 21 were amplified by PCR using primary and secondary PCR primer pairs. PCR products were purified and submitted for DNA sequencing against forward and reverse primers (analyzed with Sequencher). Results: Mean baseline plasma VEGF level was 107.6 pg/ml (range, 52–198 pg/ml). For analysis, we combined CR+PR+SD versus PD. Wilcoxon rank-sum test was used to compare baseline plasma VEGF levels, urine VEGF and serum VEGFR2 between the two response groups. There were no significant differences between response and baseline VEGF levels (p=0.39), urine VEGF (p=0.56), and VEGFR2 (p=0.56) respectively. The pt with prolonged CR (18 mos) had the highest baseline VEGF level. Cox proportional hazards regression models was used to look at the association between overall survival (OS) and VEGF levels. There was no significant relationship between (OS) and baseline plasma VEGF (p= 0.38), urine VEGF (p=0.33) or serum VEGFR2 (p=0.63) respectively. No EGFR mutations were found in exon 19 (n=4) or exon 21 (n=6). EGFR expression is being evaluated. Conclusions: Our results indicate that there were no significant relationship between response or OS and baseline VEGF or VEGFR2 levels in our treated pts. The study was supported by NCI Grant N01-CM-17102. [Table: see text]