Predicting Survival in Bevacizumab-Treated Colorectal Cancer: Personalized Mathematical Models Based on Clinical and Angiogenic Biomarkers

TL;DR

This study uses biomarkers and clinical data to predict survival in colorectal cancer patients treated with bevacizumab, improving personalized treatment strategies.

Contribution

The study introduces a novel mathematical model combining clinical and angiogenic biomarkers to predict survival in bevacizumab-treated CRC patients.

Findings

Low VEGF-A and VEGF-D levels with high bFGF correlate with improved overall survival.

A logistic regression model incorporating biomarkers and clinical parameters shows significant prognostic accuracy.

Adding CypA to the model further refines survival prediction for CRC patients.

Abstract

Aberrant activation of proangiogenic signaling pathways, particularly the vascular endothelial growth factor (VEGF) axis, drives neovascularization and tumor progression in colorectal cancer (CRC). Bevacizumab targets VEGF-A-mediated angiogenesis, but the lack of validated predictive biomarkers limits personalized treatment. In this prospective study, we evaluated a panel of circulating angiogenic biomarkers combined with clinical parameters, using mathematical models to predict survival in metastatic CRC patients treated with bevacizumab and chemotherapy. Low VEGF-A and VEGF-D levels, together with high bFGF, were associated with improved overall survival (OS). A logistic regression model incorporating these biomarkers, regional lymph node invasion, and primary tumor resection status showed significant prognostic accuracy (p < 0.001). Incorporating CypA further refined the model, identifying patients with low VEGF-A, VEGF-D, and CypA, and high VEGF-C and PlGF, as having the most favorable OS. These findings demonstrate that integrating clinical and circulating biomarker data can improve individualized risk assessment and support personalized therapeutic strategies for CRC patients receiving bevacizumab.

Figures