The expression of vascular endothelial growth factor correlates with mutant p53 and poor prognosis in human breast cancer

Abstract

Wild-type p53 protein has been shown to inhibit angiogenesis through thrombospondin in the preclinical setting. Here, we determined the associations between the expression of the angiogenic factor vascular endothelial growth factor (VEGF) and the p53 status, including different mutation sites and types, in primary breast cancer. Cytosols from 224 primary breast cancer patients were analyzed with an enzyme immunoassay for determination of human VEGF165 protein content. p53 status was determined by cDNA-based sequencing of the entire coding region, by immunohistochemistry (IHC), and by a p53 luminometric immunoassay (LIA) method. Statistically significant associations was found between higher VEGF content and non-wild-type p53 status for all methods; sequence-based data (P = 0.0019), IHC data (P = 0.0068), and the LIA method (r = 0.427; P > 0.001). Highest VEGF values were detected in tumors with p53 insertions, deletions, and stop codon mutations (P = 0.0043). Combining p53 status and VEGF content resulted in additional prognostic information, relapse-free survival (RFS; P = 0.0377), overall survival (OS; P = 0.0319), and breast cancer corrected survival (BCCS; P = 0.0292). In multivariate analysis, the relative hazard increased when the VEGF data were added to the p53 status, with a relative hazard of 1.7 for RFS and 3.0 for BCCS, compared with 1.1 for RFS and 1.4 for BCCS among the patients with either high VEGF content or p53 mutation. Higher VEGF content was statistically significantly correlated with a worse outcome for patients with estrogen receptor-positive tumors receiving adjuvant tamoxifen: RFS (P = 0.0471), OS (P = 0.0134), BCCS (P = 0.0064), as well as in multivariate analysis with point estimates of 3.4 and 2.1 for BCCS and RFS, respectively. VEGF expression is related to p53 status in human breast cancer patients. Combining VEGF with p53 status resulted in better prognostic prediction.