Review of novel liquid-based biomarkers for prostate cancer: towards personalised and targeted medicine

Abstract

Background:

Prostate cancer is the most commonly diagnosed cancer in men and it is responsible for about 10% of all cancer mortalities in both American and Canadian men. At present, serum prostate-specific antigen levels remain the most commonly used test to detect prostate cancer, and the standard and definitive diagnosis of the disease is via prostate biopsy. Conventional tissue biopsies are usually invasive, expensive, painful, time-consuming, and unsuitable for screening and need to be consistently evaluated by expert pathologists and have limited repeatability. Consequently, liquid biopsies are emerging as a favourable alternative to conventional tissue biopsies, providing a non-invasive and cost-effective approach for screening, diagnosis, treatment and monitoring of prostate cancer patients.

Materials and methods:

We searched several databases from August to December 2020 for relevant studies published in English between 2000 and 2020 and reporting on liquid-based biomarkers available in detectable quantities in patient bodily fluid samples. In this narrative review paper, we describe seven novel and promising liquid-based biomarkers that potentially account for individual patient variability as well as used in disease risk assessment, screening for early disease detection and diagnosis, identification of patients’ risk for metastatic disease and subsequent relapse, monitoring patient response to specific treatment and providing clinicians the potential to stratify patients likely to benefit from a particular treatment.

Conclusions:

The concept of precision medicine from prevention to treatment techniques that take individual patient variability into account will depend on the development of effective clinical biomarkers that interrogate key aberrant pathways potentially targetable with molecular targets or immunologic therapies. Liquid-based biomarkers with high sensitivity and specificity for prostate cancer are emerging as minimally invasive, lower risk, readily obtainable and easily repeatable technique for screening for early disease detection and diagnosis, patient stratification at diagnosis into different risk categories, identification of patients’ risk for metastatic disease and subsequent relapse, and real-time monitoring of patient response to specific treatment. Thus, effective liquid-based biomarkers will potentially shift the treatment paradigm of prostate cancer towards more personalised and targeted medicine.

Introduction

Prostate cancer is the most commonly diagnosed cancer in men and it is responsible for about 10% of all cancer mortalities in both American and Canadian men. In 2020, it was estimated that approximately 23,300 new prostate cancer cases will be diagnosed and more than 4,200 men will die from the disease in Canada ^1,2 and about 238,590 newly diagnosed cases and 29,720 mortalities in the United States. ³ At present, serum prostate-specific antigen (PSA) levels remain the most commonly used test to detect prostate cancer, and the standard and definitive diagnosis of the disease is via prostate biopsy. However, several studies ^4–23 have reported that the serum PSA level can also be elevated in benign conditions, has limited specificity, and results in high rates of over-diagnosis and treatment of indolent prostate cancers. Furthermore, it has a low positive predictive value and results in a significant proportion of negative biopsies which often leads to repeat PSA measurements and biopsies. Reed and Parekh ¹⁸ reported a negative biopsy rate of approximately 60–75% in men with serum PSA levels in the 3–10 ng/mL range, Hu et al. ¹⁹ also reported that only 25% of patients with PSA levels in the 2–4 ng/mL range have prostate cancer, and according to White et al., ²⁰ approximately 85% of men with PSA levels <4 ng/mL show non-cancerous biopsy results. Additionally, screening and diagnosis of prostate cancer has been hampered by the inability to predict who has the potential to develop fatal disease and who has indolent cancer. Consequently, several studies have focused on the identification of effective prostate cancer-specific biomarkers that can be utilised for early detection of disease, prediction of disease aggressiveness and as therapeutic targets for androgen-independent prostate cancer. ^{6–21,24–96} The identification of effective clinical biomarkers that potentially accounts for individual patient variability and capable of interrogating key genetic aberrant pathways likely targetable with molecular targets or immunologic therapies will likely shift the diagnoses and treatment of prostate cancer towards more personalised and targeted medicine.

Although biomarkers commonly used for diagnosis, prognostication and therapeutic monitoring can be obtained through either liquid or tissue biopsies, liquid biopsies are emerging as a favourable alternative to conventional tissue biopsies by providing a cost-effective non-invasive approach for the detection and monitoring of cancer. Conventional tissue biopsies are typically surgical biopsy, radiologically guided biopsy or endoscopic biopsy and are invasive, expensive, painful, difficult to obtain, time-consuming, not suitable for screening, required to be consistently evaluated by expert pathologists, limited in repeatability, and limited by patient comorbidity and may introduce clinical risks to patients. ^22,23 Liquid-based biomarkers on the other hand can be found in several patient body fluids, including blood serum and plasma, pleural fluid, saliva, exhaled breath condensate, and urine. Liquid biopsy offers a new and unique opportunity to identify potential tumour biomarkers and has a significant number of advantages such as being relatively less invasive, minimally painful, lower risk, inexpensive, readily obtainable and easily repeatable as well as has the potential for screening and real-time monitoring of patient treatment. ^{22,23,97–100} In a previous study, Osei and Swanson ⁴ presented a comprehensive literature review of currently available clinical biomarkers, and more recently Osei et al. ⁵ reported on novel and promising tissue-based biomarkers for the management of prostate cancer. The goal of this narrative review paper is to describe seven novel and promising liquid-based biomarkers that could potentially account for individual patient variabilities and capable for disease risk assessment, screening for early detection and diagnosis, identification of patients’ risk for metastatic disease and subsequent relapse, and monitoring patient response to specific treatment and could provide clinicians the potential to stratify patients likely to benefit from a particular treatment.

Materials and Methods

The following databases PubMed, PMC, NCBI, PNAS, Springer Link, Wiley Online Library, Lacent, Science Direct and Medline were searched from August to December 2020 for relevant studies published in English between 2000 and 2020 and reporting on biomarkers available in detectable quantities from bodily fluid samples and has the potential for diagnostic, predictive, prognostic and therapeutic monitoring of prostate cancer patients. The literature search used the following keywords: ‘emerging biomarkers and prostate cancer’, ‘prostate cancer biomarkers’, ‘liquid biomarkers for prostate cancer’, ‘liquid-based biomarkers’, ‘biomarkers for prostate cancer’ and novel biomarkers and prostate cancer’. The searches were not limited by study design and included conference abstracts, full research articles and reviews. Several publications on biomarkers for the management of prostate cancer were identified, and we excluded articles that report on currently available clinical biomarkers and novel tissue-based biomarkers. For the current report, we reviewed articles reporting on novel and promising liquid-based biomarkers for prostate cancer management.

Ephrin-A2

Fourteen human type 1 transmembrane protein members of the erythropoietin-producing hepatoma (Eph) family of receptors make up the largest family of receptor tyrosine kinases and are associated with eight members of a class of ligands known as Ephrins. ^24–33 Fourteen Eph receptors are divided into the Eph-A subclass of nine members (Eph-A1 to Eph-A8, and Eph-A10) and the Eph-B subclass of five members (Eph-B1 to Eph-B4, and Eph-B6) based on amino acid sequence homology and relative binding affinities to glycosylphosphatidylinositol-linked Ephrin-A or transmembrane Ephrin-B ligands. ^{24–26,28–31} Similarly, the eight Ephrin ligands are also divided into the Ephrin-A subclass of five members (Ephrin-A1 to Ephrin-A5) and the Ephrin-B subclass of three members (Ephrin-B1 to Ephrin-B3). ^25,29–31 Numerous studies have reported that several Eph receptors and their Ephrin ligands regulate contact inhibition of locomotion, control cell movements by regulating the actin cytoskeleton, are activated by cell–cell contact and are involved in the development or progression of certain cancers including prostate. ^{24,26,29,30,32,33} The Ephrin-A2 (EFNA2) gene is located on the short arm of chromosome 19 at position 13.3 (19p13.3) and encodes a membrane-bound protein with a glycosylphosphatidylinositol lipid anchor which is a ligand to several Eph receptors including Eph-A3 to Eph-A5 and Eph-A8. The Ephrin-A2 is implicated in several physiological processes including tissue and organ development, neovascularisation, cell adhesion and separation, as well as the progression of many diseases including oncogenesis. ^{24,26,27,29,30} Li et al. ²⁷ reported that the level of circulating Ephrin-A2 in serum is significantly elevated in prostate cancer patients compared to benign prostatic hyperplasia (BPH) patients and is also up-regulated in patients with osteoporosis, which is a common complication of prostate cancer patients undergoing androgen deprivation therapy. Several studies have reported that the Ephs–Ephrins signalling plays a key role in developmental processes, including axon guidance, cancers of epithelial origin through stimulation of oncogenic transformation, tumour angiogenesis, tissue boundary formation, and promotion of increased cell survival, invasion and migration. ^{24–30,32,34,35} Furthermore, serum Ephrin-A2 is reported to be a potential diagnostic and prognostic biomarker for the management of targeted prostate cancer therapy. ^27,36

Li et al. ²⁷ investigated whole serum Ephrin-A2 and circulating exosomal Ephrin-A2 in clinical samples from 20 healthy controls, 21 BPH patients and 50 prostate cancer patients. They also assessed any correlation between exosomal Ephrin-A2 and tumour–node–metastasis (TNM) staging and Gleason score as well as its diagnostic effectiveness for BPH and prostate cancer. They observed significantly elevated expression levels of whole serum Ephrin-A2 and exosomal Ephrin-A2 in prostate cancer patients compared to BPH patients and controls. They reported that exosomal Ephrin-A2 expression is positively correlated with TNM staging and Gleason score and has superior diagnostic efficiency over serum PSA in discriminating prostate cancer patients from BPH patents. They concluded that exosomal Ephrin-A2 has high potential as a biomarker for the detection of prostate cancer and offers a new therapeutic target for the disease. In another study, Li et al [36] evaluated the diagnostic and prognostic importance of Ephrin-A2 levels in 88 serum samples and reported mean serum circulating Ephrin-A2 levels of 522 ± 36 ng/L, 394 ± 38 ng/L and 257 ± 46 ng/L in prostate cancer patients, BPH patients and controls, respectively. Furthermore, they reported significantly higher serum circulating Ephrin-A2 levels in patients with a high Gleason score of 8–10 compared to patients with low Gleason score of 6–7. They concluded that low serum Ephrin-A2 levels are associated with the presence of BPH, while elevated levels are associated with the presence of prostate cancer, and serum Ephrin-A2 level can differentiate prostate cancer from BPH more effectively than pre-surgical serum PSA level, making it a potential diagnostic and prognostic biomarker, as well as a promising molecular therapeutic target to mitigate prostate cancer progression. ³⁶

Glutathione S-transferase pi 1 (GSTP1)

The glutathione S-transferases (GSTs) are a family of phase 2 detoxification enzymes that catalyse the conjugation of glutathione to a broad variety of endogenous and exogenous electrophilic compounds. They are divided into two distinct super-family members, namely the membrane-bound microsomal GSTs, which play a key role in the endogenous metabolism of leukotrienes and prostaglandins, and the cytosolic GSTs which are highly polymorphic and are divided into six classes (i.e., α, κ, μ, π, σ and θ). ^37–42 The GST pi 1 (GSTP1) gene is located on the long arm of chromosome 11 at position 13.2 (11q13.2) and encodes the GSTP1 protein which is implicated in a large number of detoxification and metabolic reactions, as well as prevention of cell genome damage and carcinoma. ^37–40,42 According to Schnekenburger et al. ³⁸ and Mahon et al., ⁴⁰ GSTP1 is involved in cell death regulation as well as interacts with apoptotic signalling pathways, and the epigenetic silencing of the gene is reported as the most common (>90%) genetic alteration in prostate cancer. The loss of GSTP1 expression as a consequence of gene methylation is reported to be associated with the first events to cause a preneoplastic phenotype to develop into a malignant phenotype ⁴² and also increases the sensitivity to metabolic or environmental toxins as well as promotes mutations and cancer development. ³⁸ Several studies have reported that the hypermethylation of GSTP1 in prostate cancer patients may lead to inactivation of GSTP1 expression, is frequently associated with tumour development, progression or poor prognosis, and is confined to prostate cancer cells as well as prostatic intraepithelial neoplasia. ^{38–40,42,43} According to Martignano et al., ⁴² approximately 70–80% of prostate cancer cases are hypermethylated whereas BPH cases are normally hypomethylated; thus, GSTP1 expression has the potential to discriminate benign from malignant transformations. Consequently, epigenetic silencing of GSTP1 could be a potential marker of the transformation from normal prostate epithelium to prostate carcinoma. Mahon et al., ⁴⁰ reported that serum-free methylated GSTP1 (mGSTP1) in circulation is associated with tumour aggressiveness and disease burden and is more specific to prostate carcinoma since it is undetected in normal prostate tissue. Therefore, according to Schnekenburger et al., ³⁸ the detection of GSTP1 methylation in all types of body fluids of prostate cancer patients represents a promising epigenetic biomarker for early prostate cancer diagnosis.

Mahon et al. ⁴⁰ investigated the relationship between serum-free mGSTP1 and treatment outcomes in 600 patients enrolled in a phase 3 multicentre randomised clinical trial assessing the impact of concurrent use of custirsen and docetaxel for the treatment of metastatic castration-resistant prostate cancer (mCRPC). They evaluated the relationships between serum-free mGSTP1 at baseline, change in mGSTP1 after docetaxel treatment, overall survival and time to PSA progression. They observed detectable levels of mGSTP1 at baseline in over 81% of patients and reported that undetectable levels of mGSTP1 at baseline and post-treatment are significantly associated with longer overall survival and longer time to PSA progression. They concluded that serum-free mGSTP1 deoxyribonucleic acid (DNA) is an early response and prognostic marker in mCRPC patients receiving first-line docetaxel treatment and has potential usefulness in the clinic to guide treatment decisions. Bastian et al. ⁴³ studied circulating cell-free DNA GSTP1 CpG island hypermethylation in preoperative serum from 85 clinically confirmed gland-confined prostate cancer patients treated with radical prostatectomy, 35 patients with negative biopsy and a dataset of 55 paired samples from radical prostatectomy treated patients with gland-confined disease and matched for Gleason score (i.e., 55 patients with PSA recurrence and 55 patients who were free of disease at last follow-up). They reported undetectable levels of circulating cell-free DNA with GSTP1 CpG island hypermethylation in serum of patients with negative biopsy; however, 12% of patients with localised disease and 28% with metastatic disease showed detectable levels. Furthermore, in the matched dataset, they reported GSTP1 CpG island hypermethylation positivity in 15% (8 out of 55) of patients who developed PSA recurrence and GSTP1 CpG island hypermethylation negativity in all disease-free patients, and serum DNA with GTSP1 CpG island hypermethylation positivity is a significant predictor of PSA recurrence. They concluded that GSTP1 CpG island hypermethylation is an important DNA-based prognostic serum biomarker for prostate cancer. Zhou et al. ³⁹ conducted a comprehensive systematic review and meta-analysis of 15 studies involving 1540 samples to investigate the association between GSTP1 promoter methylation and incidence of prostate cancer. They reported a high incidence of GSTP1 promoter methylation in patients with prostate cancer compared to those without the disease and observed that GSTP1 promoter methylation is highly correlated with prostate cancer incidence. They concluded that methylation of the GSTP1 promoter increases the risk of prostate cancer and is a potential biomarker for disease diagnosis.

Serum metabolites

Metabolites are small molecule intermediates or end products (e.g., amino acids, organic acids and bases, fatty acids, bile acids, lipids, and carbohydrates) of enzyme-catalysed metabolic reactions within a cell that characterises the cell’s metabolism in physiological or pathophysiological conditions and are altered in diseases such as carcinomas. ^44–49 Chaleckis et al. ⁴⁸ reported that metabolites in human blood can provide evidence for individual physiological states influenced by genetic, epigenetic and lifestyle factors. Dysregulation of metabolism has been reported to play an important role in the development and progression of prostate carcinoma, and in addition, differences in serum or plasma metabolite concentrations and prostatic secretions between patients and healthy controls have also been reported. ^{44–47,50–53} According to Lima et al., ⁴⁵ neoplastic cells have a unique metabolic phenotype associated with cancer development and progression, and therefore, the identification of dysfunctional metabolic pathways via metabolomics has the potential to identify new diagnostic markers for the detection and prognosis of diseases, to monitor the response to therapeutic interventions and to facilitate the discovery of therapeutic targets. Schmidt et al. ⁵⁴ and Trock ⁴⁴ have reported various metabolites that are associated with the risk of prostate cancer, advanced stage disease, disease aggressiveness and prostate cancer-specific mortality. Röhnisch et al. ⁵² have also reported an association between pre-diagnostic levels of plasma and serum metabolites and the risk of prostate cancer incidence. According to Trock, ⁴⁴ due to the functional significance of metabolites, metabolomic biomarkers or profiles hold promise for personalised and targeted medicine.

Giskeødegård et al. ⁵⁰ investigated differences in metabolic markers in serum and plasma samples from 29 prostate cancer patients and 21 controls with BPH via metabolomic analysis using magnetic resonance spectroscopy, mass spectrometry and gas chromatography. They reported that metabolic profiles have a potential to discriminate prostate cancer patients from controls with BPH with a sensitivity and specificity of 82% and 75%, respectively. They observed significantly different levels of metabolites such as amino acids, lipids and metabolites involved in energy metabolism between patients and controls. They also reported elevated levels of decanoylcarnitine, tetradecenoylcarnitine, octanoylcarnitine, dimethyl-sulfone, phenylalanine and lysine metabolites in prostate cancer patients, whereas phosphatidylcholine diacyl and lipid2 metabolites levels were elevated in the controls. They concluded that a combined analysis of serum and plasma samples by different metabolomic measurement techniques has the potential to discriminate prostate cancer from controls, and that changes in fatty acid, choline and amino acid metabolism are potential markers for prostate cancer. Stabler et al. ⁵⁵ analysed serum samples from 58 patients with rapid recurrent prostate cancer and recurrence-free patients after radical prostatectomy for metabolites capable of differentiating patients who developed early biochemical recurrence within 2 years of surgery from those who remained recurrence-free after more than 5 years. They reported elevated concentrations of homocysteine, cystathionine and cysteine in the serum of patients with recurrent disease compared to recurrence-free patients. They concluded that elevated levels of serum homocysteine, cystathionine and cysteine concentrations independently predict both risk of early biochemical recurrence and disease aggressiveness.

Zang et al. ⁵¹ studied untargeted metabolomic profiling of age-matched serum samples from 64 prostate cancer patients and 50 healthy individuals to profile and identify a panel of metabolites in blood serum that discriminates cancer patients from healthy individuals. They used ultraperformance liquid chromatography coupled to high-resolution mass spectrometry and tandem mass spectrometry combined with machine learning techniques. They reported a panel of 40 metabolic spectral features capable of discriminating prostate cancer patients from controls with 92% sensitivity, 94% specificity and 93% accuracy. Furthermore, they reported that the identification of fatty acids, amino acids, lysophospholipids and bile acids provided additional insights into the metabolic alterations associated with the disease. Röhnisch et al. ⁵² investigated the association between plasma metabolite concentrations and the risk of prostate cancer in 777 cancer patients and 777 matched controls stratified by disease aggressiveness and age at baseline. They reported that lysophosphatidylcholines are positively associated with the risk of overall prostate cancer with stronger correlation observed in older patients, and glycine positivity is associated with the risk of overall prostate cancer in younger patients. They concluded that several glycerophospholipids are positively associated with the risk of overall and aggressive prostate cancer. Huang et al. ⁵³ investigated metabolites in pre-diagnostic serum from 197 prostate cancer cases in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) Study and identified 625 metabolites. They reported that sterol/steroid metabolite score increased the risk of prostate cancer death by 46%, and lipids and elevated levels of serum N-oleoyltaurine are significantly associated with prostate cancer-specific mortality. They concluded that pre-diagnostic serum N-oleoyl taurine and sterol/steroid metabolites are associated with prostate cancer survival.

Ferritin

Ferritin is an intracellular protein that stores iron and releases it in a controlled manner, thereby acting as a buffer against iron deficiency and iron overload in humans, and is present in both glycosylated and non-glycosylated forms. ^56–60 According to Wang et al. ⁵⁸ and Su et al., ⁵⁹ ferritin plays important roles in many physiological and pathological processes, including immunosuppression, neurodegenerative, cardiovascular diseases, proliferation, angiogenesis and carcinogenesis. In general, the normal serum ferritin levels in men range 20–250 µg/L; however, these levels are elevated in patients with a variety of malignant cancer types and are correlated with the risk of prostate cancer. ⁵⁸ According to Vela, ⁶⁰ prostate cancer cells require iron for survival as it is used in the activity of enzymes that control androgen receptor transcriptional activity, which is a known promoter of the disease. Several studies have reported that elevated levels of serum ferritin are associated with iron overload, inflammatory responses such as cancer, infections, autoimmune diseases and poor prognosis for prostate cancer. ^57,59,60 According to Zhao et al., ⁵⁷ serum ferritin level is a potential diagnostic indicator of prostate cancer and others ^59,61,62 have also reported the increasing use of urine as a source of biomarkers for prostate cancer screening due to its non-invasiveness, convenience, easily available and potential to capture the disease at an earlier stage. Su et al. ⁵⁹ reported that the changes of the composition, quantity and quality of urine can provide information that reflect the generation, development and prognosis of urinary diseases including prostate cancer, and thus an increasing number of investigations are ongoing in urinary proteomics in prostate cancer biomarkers research.

Studies have reported an association between high serum or urinary ferritin levels and an increased risk of prostate cancer, although some of the results are inconsistent. ^{57–59,63,64} Kuvibidila et al. ⁶³ evaluated ferritin levels in serum samples from 34 newly diagnosed, untreated prostate cancer patients and 84 healthy individuals ranging in age from 49 to 78 years to determine whether elevated serum ferritin (i.e., > 300 μg/L) is associated with increased risk of prostate cancer. They reported significantly lower mean concentrations of serum ferritin in prostate cancer patients and concluded that serum ferritin levels are negatively associated with the presence of prostate cancer. On the contrary, Wang et al. ⁵⁸ conducted a large-scale case–control study consisting of 2002 prostate cancer patients and 951 BPH patients as control to examine the relationship between serum ferritin and the risk, diagnosis and prognosis of prostate cancer. They reported that elevated levels of serum ferritin are positively associated with increased PSA levels, prostate cancer risk, higher Gleason scores and higher predictive accuracy for serum PSA. They concluded that serum ferritin is significantly associated with prostate cancer risk and is a potential non-invasive biomarker to complement PSA test in the diagnosis and prognostic evaluation of the disease. Zhao et al. ⁵⁷ also investigated the potential of urinary ferritin as a prostate cancer diagnostic biomarker in three prostate cancer patients, three BPH patients and three age-, gender-, and ethnicity-matched healthy individuals. They reported 15 overexpressed proteins including ferritin heavy-chain gene and ferritin light-chain gene in the urine of prostate cancer patients compared to BPH patients and control individuals. They concluded that there are notable differences in urinary proteins among patients with prostate cancer, BPH and normal controls, and both ferritin heavy chain and ferritin light chain are potential biomarkers for prostate cancer that play important but different roles in the disease. Su et al. ⁵⁹ also conducted a study to identify potential urine biomarkers capable of identifying and discriminating prostate cancer from BPH in a cohort of nine prostate cancer patients, nine BPH patients, and nine sex- and ethnicity-matched healthy controls. They reported significantly elevated levels of ferritin in the urine of prostate cancer patients in comparison with BPH patients and controls and concluded that urine ferritin has the potential to distinguish prostate cancer patients from BPH patients. Solé et al. ⁶² used the next-generation sequencing technique to evaluate the urinary transcriptome of 178 whole urine samples from high-stage and low-stage prostate cancer patients as well as from BPH patients. They identified and validated five differentially expressed messenger RNA (mRNA) biomarkers (i.e., ferritin heavy chain 1 (FTH1), bromodomain and PHD finger containing 1 (BRPF1), oxysterol binding protein (OSBP), polyhomeotic homolog 3 (PHC3) and uveal autoantigen with coiled-coil domains and ankyrin repeats (UACA)) capable of discriminating prostate cancer patients from BPH patients and healthy controls. They concluded that urine is a valuable source of biomarkers that warrants further investigation for clinical utilisation.

Prostate-specific membrane antigen (PSMA)

Prostate-specific membrane antigen (PSMA) is a type II membrane glycoprotein with an intracellular segment, a transmembrane domain and an extensive extracellular domain ^5,65–74 that, in humans, is encoded by the PSMA (or FOLH1, folate hydrolase 1) gene located on the short arm of chromosome 11 at position 11.12 (11p11.12). ^71,72 The expression and enzymatic activities of PSMA are reported to be elevated in malignant prostatic epithelial cells and high-grade prostatic intraepithelial carcinoma; however, it is observed to be absent or moderately expressed in hyperplastic and benign tissue. ^{67–72,74,75} Several studies have reported that high PSMA levels correlate with tumour stage, tumour aggressiveness, androgen independence, metastatic disease, Gleason score, as well as disease recurrence, and the PSMA level is maintained in hormone-refractory prostate cancer. ^{68,70,72,74,75} This suggests that PSMA expression is an adverse prognostic factor for prostate cancer and an independent predictor of disease recurrence. ^{65,68–70,75} According to Wüstemann et al., ⁶⁷ this tumour-associated biomarker provides the possibility for the development of new strategies for diagnosis and therapy of prostate cancer, since it is known to be highly overexpressed on malignant prostate tumour cells and its expression level correlates with disease aggressiveness. Therefore, several studies are exploiting PSMA specificity to prostate cancer for diagnostic imaging strategies using anti-PSMA antibodies and its upregulation by androgen deprivation for anti-PSMA therapies for the treatment of hormone-refractory prostate cancer and metastatic disease. ^65,67–72

Paller et al. ⁶⁹ investigated PSMA expression in circulating tumour cells (CTCs) at the time of therapeutic decisions and its correlation with PSA and/or radiographic progression or response in 15 patients undergoing sequential therapies with androgen receptor-targeting agents, taxane chemotherapies and PSMA-directed therapies. They reported PSMA positivity at baseline in about 73% (11 of 15) of patients and PSMA negativity in the remaining 27% of patients. Furthermore, they reported that PSMA expression significantly decreased in patients treated with Lu-PSMA-617 (a PSMA-targeted radioligand therapy) and serum PSA levels declined with decreased PSMA expression. They concluded that PSMA expression in CTCs is a dynamic marker, PSMA transcript decline is associated with concurrent decreases in serum PSA, and sequential CTC sampling could provide a non-invasive response assessment to systemic treatment for mCRPC. Zhang et al. ⁶⁶ evaluated the use of the MJ Opticon Monitor II apparatus for real-time quantitative polymerase chain reaction (PCR) of PSMA in peripheral blood specimens from 13 healthy individuals, 17 metastatic prostate cancer patients, 20 patients with localised prostate cancer and 68 patients with BPH. They reported significant differences in PSMA mRNA levels in blood samples among BPH, locally confined prostate cancer and metastatic prostate cancer and demonstrated that real-time quantitative PCR is a sensitive, accurate and high reproducibility technique for PSMA detection in peripheral blood. They concluded that the technique improves accuracy and reliability of assessment of disseminated prostate cancer cells in peripheral blood and is promising for clinical diagnosis purposes. Xiao et al. ⁷³ used a novel protein biochip immunoassay technique to quantify and compare serum PSMA levels in healthy men and patients with either benign or malignant prostate disease. They reported mean serum PSMA levels of 623 ng/mL, 117·1 ng/mL, 272·9 ng/mL and 359·4 ng/mL for prostate cancer patients, BPH patients, control group aged <50 years and control group aged >50 years, respectively. They concluded that serum PSMA may be a more effective marker than PSA and a clinically useful diagnostic biomarker to improve the specificity in differentiating prostate cancer from BPH and thus warrants further study to evaluate its clinical utility. Nagaya et al. ⁷⁴ investigated the clinical significance of PSMA expression in 203 CTC samples from 79 CRPC patients and further evaluated the association between PSMA expression in CTCs and treatment response in 71% (i.e., 56 out of 79) of the patients who progressed on therapy and were subsequently provided a new treatment. They reported that 36% (i.e., 20 out of the 56) of the patients expressed PSMA positivity in CTCs and also PSMA expression is inversely associated with percentage change in PSA. Furthermore, they reported that PSA progression-free survival and overall survival are significantly shorter in the PSMA-positive patients. They concluded that PSMA expression is predictive of poorer treatment response, shorter PSA progression-free survival and overall survival, and PSMA expression in CTCs is a novel poor prognostic marker for CRPC patients.

MicroRNA (miRNA)

MicroRNAs (miRNAs), first discovered in 1993, are a class of endogenous small non-coding ribonucleic acid (RNA) molecules (19–22 bases in length) that regulate protein-coding gene expression by repressing translation or cleaving RNA transcripts in a sequence-specific manner. ^76–79 Several studies have reported that miRNAs regulate the expression of several genes and thus are involved in almost all biological processes, including genome stability, cell development, proliferation, cell cycle, apoptosis, adhesion, migration, inflammation, invasion and metastasis ^76–78 According to Goto et al., ⁷⁶ miRNAs regulate the expression of approximately 60% of all genes and thus are considered as a mechanism of fine-tuning regulation. Liu et al. ⁷⁸ reported that miRNAs affect tumorigenesis mainly by interrupting the cell cycle at the cellular level and by interacting with signalling pathways, oncogenes and with the response to environmental factors at the molecular level. Aberrant miRNA expression is reported to contribute to cancer initiation, progression, metastasis and drug resistance by targeting several cancer-related genes moreover, highly expressed miRNAs function as oncogenes by repressing tumour suppressors and low miRNA expression suppresses tumours by negatively regulating oncogenes. ^{76,78,80–85} Studies have reported that over half of miRNA genes are located in or near fragile sites or cancer-associated genomic regions which are preferential sites of sister chromatid exchange, deletion, translocation, amplification or integration of plasmid DNA and tumour-associated virus, which frequently cause aberrant miRNA expression during pathogenesis. ^76–78 According to Goto et al. ⁷⁶ and Massillo et al., ⁷⁷ there is a growing body of evidence suggesting that miRNAs contribute to prostate cancer development, progression and metastasis. Massillo et al. ⁷⁷ and Bidarra et al. ⁷⁹ also reported that miRNAs are observed to circulate in biological fluids including serum with significant stability, thereby making circulating miRNAs potential tumour-specific and fluid-circulating biomarkers for diagnosis and follow-up of prostate cancer patients. According to Massillo et al., ⁷⁷ miRNA expression could predict clinical outcome as well as prognosis and could be used as markers of disease relapse and metastases. Therefore, miRNAs are emerging as promising tools for the diagnosis, prognosis and management of prostate cancer patients as well as for potential miRNA-based cancer treatment.

Al-Kafaji et al. ⁸⁶ assessed the potential use of circulating miRNAs in serum to distinguish localised prostate cancer from BPH and to discriminate between low- and high-risk patients at an early stage in a population of 35 patients with localised disease, 35 BPH patients and 30 healthy individuals. They stratified patients based on tumour stage, PSA level and Gleason score into low-risk (T1 or T2, PSA < 10 ng/ml or GS ≤ 7) and high-risk (T3 or T4, PSA > 20 ng/mL or GS ≥ 8) cohorts. They reported that patients with localised prostate cancer exhibited significantly lower expression of miR-15a, miR-126, miR-192 and miR-377 compared to patients with BPH or healthy subjects. Furthermore, they reported that the expression of the four miRNAs was lower in high-risk prostate cancer patients compared to low-risk patients and was significantly and independently associated with prostate cancer risk. They demonstrated that the miRNAs significantly distinguished prostate cancer patients from BPH patients and controls, and low-risk prostate cancer patients from high-risk cancer patients. They concluded that the expression of these miRNAs in circulating blood is a promising, non-invasive biomarker for the early detection of localised prostate cancer, and for prostate cancer risk stratification; however, they suggested validations of the clinical implementation in a larger cohort. Hoey et al. ⁸⁷ investigated the expression levels of circulating miRNAs and their potential to independently predict patient risk stratification after radical prostatectomy in 78 patients. They stratified patients into high-risk and low-risk categories based on Gleason score, pathological T stage, surgical margin status and diagnostic PSA level. According to their categorisation, all patients with Gleason score≥8, pT3a and positive margin, pT3b and any margin, or diagnostic PSA > 20 μg/mL were classified as high risk (n = 44) and all other patients were classified as low risk (n = 31). They identified four miRNA (miR-17, miR-20a, miR-20b and miR-106a) that are capable of discriminating high-risk from low-risk patients and reported that elevated expression levels of these miRNAs are associated with shorter time to biochemical recurrence and confer an aggressive phenotype. They concluded that these circulating miRNAs have the potential to independently predict risk stratification of prostate cancer patients after radical prostatectomy.

Huang et al. ⁸⁸ studied circulating plasma exosomal miRNAs that could be associated with overall survival in a screening cohort of 23 CRPC patients and identified miRNAs were further evaluated for prognosis in a follow-up cohort of 100 patients. They observed an association of miR-1290, miR-1246 and miR-375 with overall survival and reported that higher levels of miR-1290 and miR-375 are significantly associated with poor overall survival in the follow-up cohort. Furthermore, they reported that the combination of miR-1290 and miR-375 into accepted clinical prognostic factors-based models in the CRPC stage significantly improved the predictive performance. They concluded that plasma exosomal miR-1290 and miR-375 are promising prognostic biomarkers for CRPC patients; however, they suggested that prospective validation is required for evaluation of these candidate miRNAs since the non-invasive blood-based test has great potential in the management of late-stage prostate cancer. Bidarra et al. ⁷⁹ examined the potential of circulating plasma miR-128-5p and miR-375-3p levels as non-invasive screening and prognostic biomarkers in 252 clinically localised prostate cancer patients undergoing curative-intent treatment and 52 controls. They reported significantly elevated levels of circulating miR-128-5p expression in prostate cancer patients and patients who developed disease metastasis compared to controls. Furthermore, they reported that miR-182-5p expression levels identified prostate cancer patients with a 77% specificity and 99% negative predictive value, and both circulating miR-182-5p and miR-375-3p levels are associated with more advanced pathologic stages. In addition, they indicated that at the time of diagnosis, circulating miR-375-3p levels predicted patients who would develop metastasis with approximately 50% sensitivity, 76% specificity and 89% negative predictive value. They concluded that these two circulating miRNAs could be clinically useful as non-invasive biomarkers for detection and prediction of metastasis development at diagnosis when used in conjunction with clinical variables used in routine practice.

Exosomes

Exosomes are small extracellular vesicles secreted by a variety of cells and contain constituents such as proteins, DNA, mRNA, miRNA, lipids, metabolites, as well as long non-coding RNA that can be taken up by distant cells and effectively alter their biological response. ^89–94 They are present and detectable in various human bodily fluids, such as blood, urine, saliva, semen, etc., and are emerging as a valued non-invasive source of diagnostic, prognostic and predictive cancer biomarkers. ^{89–91,93–96} Exosomes are reported to mediate intercellular communication and regulate tumour progression, immune responses, angiogenesis, migration, apoptosis, metastasis and drug resistance in prostate cancer. ^75,89,92 Pan et al. ⁹⁰ reported that exosomes derived from prostate cancer patients showed high levels of miR-21 and miR-141 (regulators of osteoclastogenesis and osteoblastogenesis), contained transforming growth factor beta (TGF-β) (inducer of the conversion from bone marrow mesenchymal stem cells to fibroblasts) and carried integrin α3 and integrin β1 (promoters of migration and invasion of epithelial cells). According to Lorenc et al., ⁸⁹ serum exosomes can serve as novel tools for various therapeutic approaches, such as drug delivery, anti-tumour therapy, pathogen vaccination, as well as immune-modulatory and regenerative therapies. Furthermore, they reported that urinary exosomes (e.g., miR-21 and miR-375) are an emerging source of biomarkers in the detection and prognosis of prostate cancer and a valuable aid in the decision-making process regarding prostate biopsy and the development of therapeutic strategies.

Øverbye et al. ⁹⁵ investigated the proteome of urinary exosomes in 16 patients and 15 healthy male specimens to identify proteins that are differentially expressed in prostate cancer patients compared to healthy controls. They observed 246 proteins differentially expressed between cancer patients and healthy controls and reported that 90% (221 out of 246) of the proteins are up-regulated in prostate cancer patients. When they narrowed the evaluation to a selected list of 37 proteins, they observed that 17 out of the 37 proteins had 100% specificity and over 60% sensitivity to discriminate prostate cancer patients from healthy controls. They also reported that the transmembrane protein 256 (TM256) has the highest sensitivity at 94%, whereas the LAMTOR proteins have very high specificity. They concluded that a combination of multiple proteins into a multi-panel test holds the promise to fully differentiate prostate cancer patients from healthy controls, and that urinary exosomes have potential in the diagnosis and clinical management of prostate cancer. Krishn et al. ⁹³ studied the alpha-v beta-3 (αvβ3) integrin in plasma-derived exosomes and its exosomal intercellular transfer in exosomes isolated from plasma samples from 70 prostate cancer patients and 14 age-matched healthy controls. They reported that exosome-enriched extracellular vesicles contain elevated levels of αvβ3 integrin and CD9 in plasma derived from prostate cancer patients compared to age-matched control. They also reported that the levels of αvβ3, CD63 and CD9 remain unaltered in exosomes-enriched extracellular vesicles isolated from the blood of prostate cancer patients treated with enzalutamide. They concluded that the detection of exosomal αvβ3 integrin in prostate cancer patients is a clinically valuable and non-invasive marker for prostate cancer management, and the potential transfer of αvβ3 integrin from exosomes-enriched extracellular vesicles to recipient cells provides a strong rationale for further investigation of the role of αvβ3 integrin in prostate cancer pathogenesis as well as its potential as a therapeutic target. Khan et al. ⁹⁶ investigated the expression of exosomal Survivin in plasma and serum from 39 prostate cancer patients, 20 BPH patients, 8 patients with advanced disease after unsuccessful chemotherapy with Taxotere and 16 healthy controls. They observed Survivin expressions in plasma exosomes from healthy controls, pre-inflammatory BPH patients and prostate cancer patients and reported significantly elevated plasma exosomes levels in cancer patients and those who had relapsed on chemotherapy compared to BPH patients and controls. They concluded that exosomal Survivin exists in the plasma of patients with newly diagnosed low-grade prostate cancer and thereby provides a rationale to further investigate its potential as an early, easily measurable biomarker for prostate cancer diagnosis and treatment management of patients with advanced disease. Gabriel et al. ⁹⁴ investigated the diagnostic potential of phosphatase and tensin homolog (PTEN) in plasma-derived exosomes and the intercellular exchange of PTEN via exosomes using blood samples from 30 pre-prostatectomy patients and 8 healthy, age-matched controls. They observed PTEN expression in all exosomes from prostate cancer patients, but no expression in exosomes from healthy controls and reported that PTEN is incorporated in the cargo of exosomes circulating in patients’ blood. They concluded that exosomal PTEN can compensate for PTEN loss in PTEN-deficient cells and has potential diagnostic value for prostate cancer.

Conclusion

The identification of clinical liquid-based biomarkers that potentially account for individual patient variabilities and capable of therapeutic targeting of specific genetically aberrant pathways which play key roles in malignant tumour formation as well as offers improved understanding of the biology responsible for prostate carcinogenesis and progression will likely shift the treatment paradigm of prostate cancer towards a more non-invasive personalised and targeted medicine. Liquid-based biomarkers with high sensitivity and specificity for prostate cancer have proven potential to provide a non-invasive and cost-effective approach for screening, early-stage disease diagnoses and patient stratification at diagnosis into different risk categories and guide the provision of optimal targeted treatment selection in clinical practice. Furthermore, they have potential roles in the development of novel effective therapies inhibiting prostate cancer progression and metastasis, and thereby substantially impacting disease mortality, increasing patient survival and potentially reducing the rates of invasive treatments by enhancing the safety of active surveillance strategies as well as providing clinicians the ability to identify patients likely to respond to specific treatments.