Does the serial measurement of known biomarkers help to identify those individuals with cancer who are at the highest risk of thrombosis?

Brown, Alison Marie (2025) Does the serial measurement of known biomarkers help to identify those individuals with cancer who are at the highest risk of thrombosis? Doctoral thesis (DClinSci), UNSPECIFIED.

Preview

Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (10MB) | Preview

Abstract

Cancer associated thrombosis (CAT) is a serious complication seen in patients with all types of cancer. CAT affects up to 20% of all patients diagnosed with cancer and is associated with poorer outcomes and an increased risk of mortality. Risk assessment models exist which try to predict those cancer individuals who are at the highest risk of thrombosis, though these are not without their limitations. The aim of this thesis was to investigate novel methods of predicting cancer-associated thrombosis. A variety of approaches were used to address this aim. First, a retrospective clinical audit was performed on patients (n = 75) with a diagnosis of pancreatic cancer, a cancer with a high prevalence of CAT. The Khorana score, the current standard risk assessment model, was assessed for the prediction of CAT. In this retrospective audit, the Khorana score was not found to predict CAT. Next, a prospective large-scale meta-analysis of published literature was performed to assess the potential of VEGF as a novel biomarker for CAT. The meta-analysis of eight studies, including 1547 patients with a diagnosis of a variety of primary site cancers, demonstrated that, whilst VEGF was found to be increased in cancer patients at the point of thrombosis, it could not be used for the prediction of CAT. Finally, a small prospective clinical trial (n = 54) was undertaken to assess the serial measurement at three sampling timepoints (baseline, 1-month and 3-months after the start of chemotherapy) of three biomarkers in participants treated at The Newcastle Upon Tyne Hospitals NHS Foundation Trust with a diagnosis of cancer of any primary site. This small prospective clinical trial determined that changes in neither D-dimers nor VEGF levels were predictive of CAT. There was however, a statistically significant difference (p = 0.0310) was seen between baseline and 1-month in soluble P-selectin levels in those who did and did not experience a thrombotic event. The findings presented in this thesis demonstrates that adaptation to risk assessment models may be required to improve the prediction of CAT in both pancreatic cancer and local geographical populations. Further work into the predictive capacity of VEGF and serial measurement of soluble P-selectin is required. Further work is required in larger patient cohorts and in specific cancer types using both VEGF and sP-selectin to elucidate any trends which could be used for the prediction of CAT.