Circulating Tumor Cells And Tumor-Derived Extracellular Vesicles As Biomarkers For Prostate Cancer
Eshwari Dathathri is a PhD student in the Department of Medical Cell Biophysics. Promotors are prof.dr. L.W.M.M. Terstappen and prof.dr. R. Bansal from the Faculty of Science & Technology.
Prostate cancer is the most prevalent cancer among men, with approximately 1.4 million diagnoses reported worldwide in the year 2020. Although most patients are diagnosed with localized disease, about 25 % are diagnosed with metastasis with poor outcomes. Despite advances in treatment options, treatment outcomes vary widely due to tumor heterogeneity in patients. This underscores the need for biomarkers to evaluate tumor heterogeneity in patients and eventually ascertain which patients are likely to respond to therapy. In recent years, liquid biopsy has emerged as an attractive alternative to tissue biopsies, offering minimally invasive means to monitor the treatment responses of patients and explore tumor heterogeneity. In this thesis, we primarily focus on circulating tumor cells (CTC) as the main biomarker, while also highlighting how the inclusion of tumor-derived extracellular vesicles (tdEV) enhances our understanding of tumor heterogeneity in metastatic prostate cancer.
We begin with a comprehensive review in Chapter 1 about liquid biopsy-based circulating biomarkers such as CTCs, circulating nucleic acids, genetic markers, extracellular vesicles (EVs), tumor-educated platelets, and secretome derived from the secreting tumor cells and tumor microenvironment in metastatic prostate cancer. Among these, the three most effective prognostic biomarkers observed in the clinical settings included CTCs, tdEVs, and prostate-specific antigen (PSA), which are examined in greater detail in the following chapters.
In Chapter 2, we examined the quality and quantity of CTCs and tdEVs obtained from the peripheral blood samples of metastatic castration-naive prostate cancer (mCNPC) and metastatic castration-resistant prostate cancer (mCRPC) patients. Using the Automated CTC Classification, Enumeration, and PhenoTyping (ACCEPT), we observed higher CTC and tdEV counts in mCRPC compared to mCNPC patients. Morphological differences between disease stages were observed, with CTC clusters and CTC exhibiting heterogeneous CK expression being more prevalent in mCRPC patients. Additionally, tdEV next to CTC showed promise as a prognostic marker, and a decrease in both counts was observed in response to 6 months of therapy in mCNPC patients.
After recognizing the added prognostic value of tdEVs, along with CTCs, we further investigated their combined effect in Chapters 3 and 4. In Chapter 3, we proposed a Blood Tumor Load (BTL) metric that integrates CTC and tdEVs into a single value ranging from 0 (low) to 1 (high) to simplify result interpretation and increase the percentage of patients from which a reliable assessment can be made. In the first case study, we successfully increased the proportion of breast cancer patients for whom HER2 expression could be reliably assessed, from 35% due to the low CTC counts, to 78% by combining CTC with tdEV. In the second case study, we demonstrated that BTL can improve prediction of favorable survival in CRPC patients, compared to CTCs and tdEVs alone. To assess therapy response, BTL warrants further investigation in a larger patient cohort with extended follow-up to determine if there is a sustained reduction in CTC/BTL.
In Chapter 4, we examined the PSMA expression of individual CTCs and tdEVs obtained from the peripheral blood of 139 mCNPC patients of the PICTURES study. The blood samples were enriched using a CellSearch CTC kit with a PSMA staining reagent, scanned using the CellTracks Analyzer II and analyzed using an artificial intelligence pipeline to identify CTCs, tdEVs and quantify their PSMA staining. Spiked culture cell samples were used to determine the threshold for PSMA positivity. We observed a strong correlation between the PSMA expression of CTCs and tdEVs, indicating that tdEVs can be useful in evaluating PSMA expression in patients with 0 or low CTC counts. The percentage of patients in whom PSMA assessment was possible improved from 27% with CTC alone to 69% with a CTC – tdEV combination. In 4 patients (4%), the Fraction PSMA-positive was observed in over 75% of the CTC – tdEV combination; in 22 patients (23%), it was between 50-75% of CTC-tdEV; in 22 patients (22%), it was between 25-50%; and in 48 patients (50%), it was observed in less than 25% of the CTC-tdEV. This relationship allows further investigation into the efficacy of PSMA-targeted therapies based on tumor PSMA expression.
Following our analysis of CTC and tdEV load in mCNPC patients, we further examined tumor heterogeneity by isolating and characterizing CTCs at a single-cell level. Chapter 5 focused on measuring PSA secretion from single CTCs to explore tumor heterogeneity in mCNPC patients. We enriched CTCs from the diagnostic leukapheresis (DLA) of 18 mCNPC patients using an in-house developed EpCAM-based immunomagnetic enrichment technique. Calcein positive and CD45 negative enriched CTCs were sorted using the flow cytometer and isolated into single cells on the VyCAP nanowell array, to collect and measure PSA secretion. We detected PSA secretion from both PSMA+ and PSMA- cells, with secretion levels ranging from 4 to 11.68 pg/cell. Importantly, we noted heterogeneity in PSA secretion within each patient reflecting intra-patient tumor variability. By genetic profiling of selective tumor clones producing different PSA levels in response to therapy, we can further gain insights into the existing limitations of PSA as a response biomarker.
Given the observed variability in PSA secretions, we aimed to investigate the contributing factors and the impact of therapies on this variability. In Chapter 6, we focused on examining the effects of androgen-targeting agents (Enzalutamide) on PSA secretions from single cells, exploring how the cell cycle affects secretion patterns, and identifying other potential biomarkers from prostate cancer cell lines with varying metastatic potential. We observed higher and heterogenous PSA secretion from LNCaP compared to other cell lines. The cell cycle influenced PSA secretion from LNCaP, with increased levels observed in the G1 phase compared to the S or G2/M phase. Furthermore, Cathepsin D and Progranulin were identified as potential biomarkers to be used alongside PSA for multiplexing and single-cell platforms.
In conclusion, this thesis highlights the significance of CTCs and tdEVs as individual biomarkers in assessing tumor heterogeneity and predicting treatment response, as well as their enhanced value when used in combination. Single cell analysis of CTCs indicated variability in PSA secretion reflecting tumor heterogeneity which can be further studied to improve the potential of PSA as a biomarker. Overall, this work contributes to optimizing the use of biomarkers for improving treatment responses in patients with metastatic prostate cancer.
