Ephrin-B2 overexpression predicts for poor prognosis and response to therapy in solid tumors

Abstract

Ephrin B2 is variably expressed on tumor cells and its blockade has been shown to inhibit angiogenesis in animal models of pancreatic, colorectal, lung and head, and neck squamous cell carcinomas. However, the implications of ephrinB2 expression in cancer patients have remained elusive. In this study, we analyzed the cancer genome atlas (TCGA) for ephrinB2 expression. We report significant correlations between EFNB2 expression, overall survival and disease-free survival in head and neck squamous cell carcinoma (HNSCC, n = 519), pancreatic adenocarcinoma (n = 186), and bladder urothelial carcinoma (n = 410). In HNSCC patients, high-EFNB2 mRNA expression was associated with tumor HPV negativity, oral cavity location, alcohol intake, higher TP53 mutation, and EGFR amplification. EphrinB2 overexpression also correlated with worse response to chemotherapy and radiotherapy. The therapeutic potential of blocking ephrinB2 was validated in HNSCC patient-derived tumor xenografts and showed significant improvement in survival and tumor growth delay. Our data shows that ephrinB2 overexpression can serve as a critical biomarker for patient prognosis and response to therapy. These results should guide design of future clinical trials exploring EphrinB2 inhibition in cancer patients. © 2016 Wiley Periodicals, Inc.

Keywords: cancer; chemotherapy; ephrin; radiation therapy.

Figure 1

EFNB2 expression is an indicator poor prognosis based on TCGA data in HNSCC (519 patients), pancreatic adenocarcinoma (186 patients), and bladder urothelial carcinoma (410 patients). Survival analysis and disease-free survival are based on average mRNA expression (high = above average; low = below average). Overall survival and disease-free survival was calculated by Kaplan–Meier method using log-rank tests for comparison. Disease-free survival (DFS) was defined as time from the date of diagnosis to the date of the last known occasion that the patient was disease-free, or the date of disease recurrence (local, regional, or distant recurrence). Death without documented recurrence was censored at the date of death. Univariate Cox proportional models were used to calculate the Hazard ratio (HR). Two-sided P-values are reported for all analyses.

Figure 2

Analysis of the HNSCC patient cohort based on EFNB2 expression. (A) Bivariate analysis of patient and tumor characteristics based on EFNB2 expression. P-values below 0.05 represent significant correlation between EFNB2 expression and the analyzed parameter. Two-sided Fisher’s exact test and the Chiq-squared test were utilized for the analysis. Classifications of alveolar ridge, buccal mucosa, lip, oral tongue, and floor of mouth were re-classified to oral cavity. Tonsil was re-classified as oropharynx. (B) Subgroup analysis of HNSCC patients based on T-stage and (C) N-stage. (D) Multivariate Cox regression analysis for OS and DFS was performed for alcohol consumption frequency, TP53, EGFR, EFNB2, and disease site.

Figure 3

High EphrinB2 expression in HNSCC patients correlates with worse response to radiotherapy (A) and chemotherapy (B). Only patients who received treatment for curative intent to the primary tumor were selected for analysis.

Figure 4

Validation of the relationship between HPV status and EFNB2 expression. Sequencing data were pooled from patient data available from JHU and the TCGA. One way analysis of variance (ANOVA) with Tukey correction was applied for statistical analysis. Three stars represent P-value <0.0001.

Figure 5

EphrinB2 protein expression in HNSCC and pancreatic adenocarcinoma patient tissue. (A) Protein expression of EphrinB2 revealed through Western blot in 12 HNSCC and 12 pancreatic patient samples. (B) Correlation of mRNA to protein expression in seven HNSCC patient samples. (C) Treatment with s-EphB4-HSA of a HNSCC patient-derived xenograft from a patient with relapsed oral cavity cancer. (D) Early time point analysis showing effect of s-EphB4-HSA on total and phosphorylated levels of sEphB4-HSA.