Prostate cancer is the most common cancer amongst men and represents a major cause of cancer-related deaths (CA Cancer J Clin 2018;68(1):7-30). Patients with metastatic prostate cancer commonly respond to androgen deprivation therapy for 12-24 months, and ultimately develop castrate-resistant prostate cancer (CRPC). The progression to CRPC occurs through restored androgen receptor (AR) activity by multiple mechanisms including AR overexpression, intracrine androgen synthesis, AR splice variants, gain of function mutations in the AR, or crosstalk with AR signaling pathway (Trends Endocrinol Metab 2010;21(5):315-324).
Understanding the pathways that drive the development of CRPC led to the development and approval of drugs such as abiraterone acetate, enzalutamide, and apalutamide, which target the AR signaling pathways (N Engl J Med 2011;364:1995-2005, N Engl J Med 2012;367:1187-1197). These therapies have dramatically improved the clinical outcomes for patients with advanced prostate cancer. However, resistance to these drugs is inevitable and represents a common and major challenge in the management of prostate cancer.
Advanced prostate cancer is a bone dominant disease that presents diagnostic and therapeutic challenges. Often there is no early indicator of clinical response since changes in a typical bone scan may take months to occur and a prostate-specific antigen (PSA) may not correlate with disease course. Additionally, there is a lack of adequate tumor tissue to study the molecular changes that occur in the later stages of disease. Alternative methods of monitoring progressive metastatic disease with circulating tumor cells (CTC) have shown promise to help guide the treatment of this group of patients.
Next-generation sequencing of CTCs gives clinicians the capacity to interrogate disease evolution and identify genomic aberrations that emerge with drug resistance, especially after multiple rounds of treatments (Cancer Discovery 2018;8:269-271). Additionally, liquid biopsies enable clinicians to easily analyze blood samples to detect cancer early, monitor recurrence, predict prognosis, and select therapy.
The History of CTCs
CTCs are an exciting non-invasive technology that correlate with patient prognosis and can help guide treatment decisions in the setting of drug resistance. CTCs demonstrate the ability of a cancer cell to detach, circulate, survive, and colonize at distant sites. In 1869, Thomas Ashworth first observed CTCs in the autopsy of a patient with metastatic cancer by comparing the tumor cells found in the blood to the tumor cells found in the primary lesion. Greater than 100 years later, the FDA cleared the use of the first analytical CTC enumeration assay, CellSearch, following three pivotal trials in metastatic breast, colorectal, and prostate cancers that demonstrate the quantification of CTCs during the course of treatment is prognostic of overall survival (Clin Cancer Res 2015;21:4992-4995).
CTCs were first reported in prostate cancer in 2004 in a study using multigene reverse transcription-PCR profiling to evaluate 37 genes in a group of 23 patients (Clin Chem 2004;50(5):826-835).
This study hypothesized that CTCs could serve as a “real-time” biopsy and showed that key genes involved in prostate cancer pathogenesis were upregulated in patients with CRPC when compared to healthy volunteers. In the control group of patients, about 13 of the 37 genes were not expressed. However, in the nine patients with metastatic CRPC, genes expressed included PSA (87%), prostate-specific membrane antigen (74%), androgen receptor (70%), human glandular kallikrein, EGFR (30%), and prostate-specific gene with homology to G protein receptor (9%). Interestingly, the number of CTCs in blood samples ranged from four to 283 in 7.5 mL of blood.
A subsequent study by Moreno and colleagues examined blood samples of 37 patients with metastatic prostate cancer for analysis of CTCs in 7.5 mL blood enriched using magnetic nanoparticles targeting the epithelial cell adhesion molecule with fluorescent labeling (Urology 2005;65(4):713-718). These samples were analyzed using multiparameter flow cytometry with nucleic acid dye positive, cytokeratin positive, and CD45 negative.
In this study, the threshold of five or more CTCs per 7.5 mL of blood was analyzed in correlation with overall survival. In the group of 37 patients, 23 (62%) had five or more CTCs, which correlated with a median overall survival of 0.7 years. In stark contrast, the group that had fewer than five CTCs had an overall survival of greater than 4 years (log-rank p=0.0002). Additionally, CTCs appeared to affect overall survival in patients with hormone refractory prostate cancer when using univariate and multivariate analyses. Other platforms of CTC detection and monitoring have greatly expanded over the years and have been extensively studied by many investigators each with its advantages and disadvantages.
Methods for Identification of CTCs
The platforms developed for CTC isolation capture differences in the physical and biologic properties between CTCs and non-tumor cells by enriching CTCs. There are two general approaches to the enrichment of CTCs: positive selection of CTC-specific cell-surface markers and negative selection, which involves leukocyte-specific cell-surface markers to removed cells from the blood leaving CTCs (Nat Rev Clin Oncol 2014;11:401-412).
The CellSearch platform uses EpCAM antibody-coated magnetic beads for capture of CTCs that subsequently identify cells as positive for CK8, CK-18, and CK-19 and negative for the common leukocyte antigen (CD45). Other assays using positive selection include AdnaTest, CTC-chip, GEDI, and MagSweeper. The AdnaTest enables molecular characterization of CTCs using RT-PCR and is able to pick up mesenchymal CTCs. The negative selection technique is used by the CTC-iChip and Microfluidic Cell Concentrator (Nat Rev Clin Oncol 2014;11:401-412). The Epic Sciences Platfrom allows for analysis of mutation or copy number variation by next-generation sequencing without cell enrichment and has been validated in a small cohort of metastatic CRPC patients (J Circ Biomark 2015;4(3):1-10). Currently, CellSearch platform is the most widely used platform in prostate cancer. While sensitivity and specificity of each of these platforms varies, further work is needed to understand the utility of each of these systems.
Potential OS Biomarker
The IMMC-38 trial was the first to prospectively evaluate the impact of the change in CTC counts on overall survival after treatment (Clin Cancer Res 2008;14:6302-6309). A total of 276 patients with metastatic CRPC treated with cytotoxic chemotherapy had the change in CTCs analyzed at the initiation and completion of chemotherapy. The conclusion from this trial was that the post-treatment CTC count of greater than five cells per 7.5 mL of whole blood was associated with shorter overall survival (6.7-9.5 months vs. 19.6-20.7 months). Additionally, patients converting from a CTC number greater than five cells per 7.5 mL (considered high risk) to fewer than five cells per 7.5 mL (considered low risk) had a corresponding improvement in overall survival (median survival time increased from 6.8 months to 21.3 months).
A subsequent phase III trial by Scher and colleagues studied 711 men with CRPC prostate cancer treated with abiraterone acetate plus prednisone versus prednisone and examined the changes in CTC count at 4, 8, and 12 weeks (J Clin Oncol 2015;33(2):1348-1355). This trial found that the 2-year survival of patients was dramatically impacted by a change in CTCs at 12 weeks following treatment with abiraterone/prednisone or prednisone. For patients with CTCs less than five (considered low risk), the 2-year overall survival was found to be 46 percent versus patients with CTCs greater than five (considered high risk) who had an overall survival of 2 percent at 2 years. Additionally, these investigators found that the combination biomarker panel of CTCs and LDH satisfied the four Prentice criteria for surrogate marker (J Clin Oncol 2015;33(2):1348-1355).
Olmos and colleagues found that patients who achieved a 30 percent decrease in CTC counts at 4, 8, and 12 weeks after treatment had improved outcomes than those who did not (Ann Oncol 2009;20:27-33). A combined analysis of the COU-AA-301 and IMMC38 trials, presented at ASCO in 2015, showed that a 30 percent fall in CTC from baseline at 4 weeks was independently associated with improved overall survival in patients treated with abiraterone acetate/prednisone or chemotherapy (J Clin Oncol 2015;33:suppl;abstr 5014). A combined analysis of 6,081 patients from five trials, including COU-AA-301, AFFIRM, ELM-PC-5, ELM-PC-4, and COMET subsequently showed the discriminatory strength of multiple endpoints (J Clin Oncol 2018;36(6):572-580). This study found that the weighted c-index for the CTC conversion response (defined as ≥ 5 CTCs at baseline, ≤ 4 at 13 weeks) was 0.79 and CTC0 (nonzero at baseline and zero at 13 weeks) was 0.81 and had the highest discriminatory power for overall survival.
Detection of Drug Resistance
A better understanding of the genomic aberrations in primary and metastatic prostate cancer could lead to the delivery of more precision-based approaches to the care for patients. Detailed molecular analyses of CTCs provide additional information than simple cell enumeration and provide opportunities to study AR transcript variants. PSA and prostate-specific membrane antigen (PSMA) are upregulated following AR activation and can serve as surrogates for AR signaling in CTCs.
Antonarakis and colleagues have shown that expression of the mRNA of AR-V7 predicts failure of secondary treatment with enzalutamide and abiraterone acetate/prednisone, but still remains sensitive to docetaxel based therapies (N Engl J Med 2014;371:1028-1038). Additionally, a point mutation in F867L in the ligand-binding domain of AR confers resistance to enzalutamide (Cancer Discovery 2013;3(9):1020-1029). Others have investigated single-cell transcriptomics that can be used to identify other alternative resistant pathways for AR directed therapies, including the WNT signaling pathway (Science 2015;349(6254):1351-1356).
The Role of Whole Exome Sequencing
The genomic characterization of cancer has focused on large-scale sequencing of primary tumors. CTCs may provide an alternative sampling source for genomic analysis, but is complicated by the rare number of tumor cells with an estimated abundance of one cell per billion normal blood cells thus showing the dire need for reliable isolation and detection in blood (Nat Biotechnol 2014;32(5):429-428).
Mapping of greater than 99.995 percent of the standard exome is possible when using a validated method suggested by Lohr and colleagues (Clin Cancer Res 2004;10:6897-6904). CTCs from two patients with metastatic prostate cancer were compared to tissue from lymph node and primary tumor. The investigators created a standardized process to generate whole exome sequencing libraries from CTCs recovered from blood, which involves cell enrichment and isolation, genomic amplification, library qualification, and census-based sequencing using MagSweeper.
In this study, 70 percent of CTC mutations were observed in the matched tissue. Additionally, 10 early trunk and 56 metastatic trunk mutations were found in the non-CTC tumor. The investigators found 90 percent and 73 percent, respectively, of these mutations in the CTC exomes. Although this is quite promising, the lack of large-scale studies using this technology makes it difficult to apply to all patients.
Limitations of CTCs
The evidence for CTCs as potential biomarkers in prostate cancer is very promising. However, there are some key limitations to the current technology that make their applications difficult for every prostate cancer patient. For example, in early stage disease, there may not be enough CTCs for early detection. This technology seems to only be useful in the setting of advanced disease, but even then, tumor cells may be isolated and difficult to analyze. As mentioned above, whole exome sequencing may provide a method to analyze these cells (Nat Biotechnol 2014;32(5):429-428).
Additionally, not all patients with advanced prostate cancer appear to have detected CTC using the CellSearch platform. This platform isolates CTCs from blood using the epithelial cell markers defined as a nucleated cell greater then 4 micromolar in diameter and lack CD45 but express CK (Int J Cancer 2014;134:2284-2293). Cells that have progressed through the epithelial-to-mesenchymal transition (EMT) may lack expression of epithelial markers (Cancer Res 2011;71:6019-6029, Int J Cancer 2014;134:2284-2293). Cancer cells that have gone through EMT are thought to be more aggressive with more metastatic potential and have loss of expression of CK8, 18, and 19 (Mol Cancer Res 2011;9:997-1007), thus making early detection of progression even more important.
There are currently efforts to attempt to capture both epithelial and mesenchymal CTCs by other companies, such as the AdnaTest, (Anticancer Res 2012;32(8):3507-3513) that may address this problem. However, this platform has not been as well-validated as the currently FDA approved CellSearch platform. Additionally, tumor heterogeneity within the same individual also proves to make the utility of this test limited. Patients who have metastatic disease that seed to different parts of the body have clonal and sub-clonal populations that when biopsied have different clonal architectures (Nature 2015;520:353-357).
In advanced prostate cancer, PSA and imaging often guide each patients' treatment decisions. However, PSA has to be interpreted carefully and does not always correlate to clinical course and survival. The use of other bloodborne markers such as CTCs has shown the potential to identify response earlier in the treatment course and help select treatment for men with advanced prostate cancer.
Encouraging data is emerging from other blood-based diagnostics, such as circulating DNA, which will further contribute to identifying subtypes of men with metastatic CRPC. As our technology evolves, the use of CTCs in combination with other prostate cancer markers will assist with therapeutic decision-making and improve the outcomes for men with advanced prostate cancer.
SAVERI BHATTACHARYA, DO, is Assistant Professor at the Sidney Kimmel Cancer Center at Thomas Jefferson University. HUSHAN YANG, PHD, is Associate Professor at the Sidney Kimmel Cancer Center at Thomas Jefferson University. CHUN WANG, MD, PHD, is Visiting Professor at the Sidney Kimmel Cancer Center at Thomas Jefferson University. LEONARD G. GOMELLA, MD, FACS, is Professor, Chair of the Department of Urology, and Director of the Sidney Kimmel Cancer Network. WILLIAM K. KELLY, DO, is Professor and Chairman of the Solid Tumor Division at Sidney Kimmel Cancer Center at Thomas Jefferson University.