Novel biomarkers for cancer detection and prognostication

Abstract

In this thesis we used a variety of approaches for biomarker discovery; in Part I we assessed whether we could identify a non-invasive surrogate markers of angiogenesis, as new vessel formation plays critical roles in the growth and metastatic spread of tumors. Moreover, many agents targeting the vascular endothelial growth factor (VEGF) pathway are currently being evaluated in clinical trials, however robust surrogate endpoints of biological activity or biomarkers for patient selection are still lacking for this assessment. Circulating endothelial cells shed from the angiogenic tumor microenvironment (Chapter 3), and endothelial progenitor cells mobilized by the tumor from the bone-marrow (Chapter 4) are evaluated as surrogate markers of angiogenesis and are studied further as marker for the assessment of optimal biological dose and surrogate endpoint during treatment with phosphokinase C-inhibitor Enzastaurin, an anti-angiogenic targeting agent in combination with gemcitabine and cisplatin based chemotherapy (Chapter 5).

In Part II we addressed the question whether extracellular nucleic acids present in the plasma of patients with cancer, could be quantified using real-time amplification for use as prognostic biomarker (reviewed in Chapter 7). These circulating nucleic acids are largely derived by necrosis or apoptosis of the tumor. Most studies have evaluated whether these circulating nucleic acids were could be used as diagnostic marker, however, as this biomarker likely reflects tumor activity we investigated whether circulating nucleic acid may serve as a prognostic biomarker. Every tumor cell that releases nucleic acids (due to necrosis or apoptosis) will have multiple gene copies of mitochondrial genes compared to genomic genes, as cells contain hundreds of mitochondria compared to one nucleus. Therefore we addressed whether mitochondrial gene amplification had elevated sensitivity and specificity as biomarker compared to genomic nucleic acids. We collected plasma from patients with advanced prostate caner and patients with benign prostate hyperplasia and assessed the strength of mitochondrial nucleic acids as prognostic biomarker compared to established prognostic variables such as prostate specific antigen (PSA) and hemoglobin count (Chapter 8), and as pan-tumor marker in patients with prostate, head and neck, kidney and colorectal cancer (Chapter 9). Due to discrepancies in the literature on applicability of circulating nucleic acids as prognostic or predictive factor, we demonstrate that methodology, sample preparation and storage strongly affect results (Chapter 10).

In Part III we use a different approach of "discovery-based research" analyzing proteomic expression patterns in serum of renal cancer patients (introduction in Chapter 11). Surface Enhanced Desorption Ionisation-Time of Flight Mass Spectrometry (SELDI-TOF MS) has resulted in a serum proteome wide search of expression signatures for cancer diagnostics in a variety of cancers such as ovarian, breast, prostate, colorectal cancer. However, promising at first proteomics is criticized because of the lack of reproducibility between institutes and research groups. Therefore we first investigated the reproducibility of SELDI-TOF MS in two Dutch institutes using identical assay procedures of two previous published studies on patients with renal cell carcinoma (Chapter 12). Next, we investigated whether we could identify a serum expression signature that could predict patient outcome prior to anti-tumor treatment, and which was an independent predictor when compared to the gold standard of survival prediction, the extended Memorial Sloan-Kettering Prognostic Factors Model for survival (Chapter 13).

This thesis is framed by a general discussion on the three chapters, by means of a debate on a number of controversies in the literature concerning this thesis (Chapter 14).