Vascular Endothelial Growth Factor (VEGF) as a Biomarker for Cancer-Associated Venous Thrombosis: A Meta-analysis

Cancer-associated thrombosis affects between 1 and 20% of all patients diagnosed with cancer and is associated with significant morbidity and a poorer prognosis. Risk assessment scores exist which include the measurement of biomarkers, and which aim to identify patients at a higher risk of developing thrombotic events, but these are poor predictors and rarely used in routine clinical practice. VEGF is a potent angiogenic factor, produced by tumour cells, and released by platelets and is essential for tumour growth and progression. It also plays a role in the promotion of thrombosis through platelet activation and adhesion, and by inducing the expression of tissue factor. Therefore, the potential of VEGF to be used as a biomarker to predict cancer-associated thrombosis requires further investigation. This study reviewed the published literature to determine whether circulating VEGF levels are associated with increased risk of venous thromboembolism in patients with cancer. PubMed and OVID databases were systematically searched according to PRISMA guidelines for relevant papers using the keywords "cancer" AND "thrombosis" AND "VEGF" up to July 2023. Inclusion and exclusion criteria were applied. Seven papers (1,528 participants) were identified and included in the meta-analysis, three of which (922 participants) measured VEGF before a thrombotic event, and the remaining four (606 participants) measured VEGF at the time of the thrombosis. Our results showed that although plasma and serum VEGF tended to be higher in those who subsequently developed thrombosis than those who did not (mean difference 70.2 pg/mL for serum, and 11.44 pg/mL for plasma VEGF, 95% CI -2.39-25.73, p  = 0.10), this was not found to be statistically significant. However, analysis of VEGF following blood sampling at the time of thrombosis showed a stronger statistically significant association between increased VEGF levels and presence of thrombosis (mean difference 117.02 pg/mL for serum, and 116.6 pg/mL for plasma VEGF, 95% CI 55.42-190.82, p  = 0.0004). Based on current studies, whilst it is increased at the time of thrombosis, VEGF is not effective as a predictive biomarker of CAT.