Prostate Cancer Antigen 3 (PCA3)

The PCA3 [formerly called differential display clone 3 (DDC3)] gene is a non-coding messenger RNA (mRNA) that is significantly overexpressed in prostate cancer cells without any correlation to age, prostatic volume or other prostatic diseases such as prostatitis. ^{Reference Bourdoumis, Papatsoris, Chrisofos, Efstathiou, Skolarikos and Deliveliotis8,Reference Cao, Lee, Ning, Handy, Wagar and Meng9,Reference Cui, Cao and Li21,Reference Auprich, Bjartell and Chun23–Reference Ye, He and Wu27} The gene is located on the long arm of chromosome 9 at position 21.2 (9q21.2) and it is embedded within intron 6 of a human homolog of the Drosophila prune (PRUNE2) gene. ^{Reference Auprich, Bjartell and Chun23,Reference Auprich, Chun and Ward24,Reference Salameh, Lee and Cardó-Vila28} The PCA3 gene was first identified and characterised by Bussemakers et al. ^{Reference Bussemakers, van Bokhoven and Verhaegh29} who reported it overexpression in cancerous tissues, low expression levels in benign prostatic tissues and undetectable levels in normal tissues or organs. Other studies have also reported that over 95% of patients with confirmed prostate cancer show elevated levels of the PCA3 gene, but low levels in normal prostatic glands or benign prostatic hypertrophy. Furthermore, that its regulation is sensitive to androgen-receptor (AR) activation which is a molecular signature of prostate cancer. ^{Reference Bourdoumis, Papatsoris, Chrisofos, Efstathiou, Skolarikos and Deliveliotis8,Reference Cao, Lee, Ning, Handy, Wagar and Meng9,Reference Auprich, Bjartell and Chun23–Reference Marks and Bostwick25,Reference Ye, He and Wu27} Overexpression of PCA3 is about 60–100% greater in cancerous cells than in benign prostate tissues thereby giving the gene a very high cancer specificity. ^{Reference Bourdoumis, Papatsoris, Chrisofos, Efstathiou, Skolarikos and Deliveliotis8,Reference Cao, Lee, Ning, Handy, Wagar and Meng9,Reference Marks and Bostwick25} It is also reported as the most specific prostate cancer biomarker identified which is capable of discriminating benign from malignant prostate cells with about 100% accuracy. ^{Reference Bourdoumis, Papatsoris, Chrisofos, Efstathiou, Skolarikos and Deliveliotis8,Reference Cao, Lee, Ning, Handy, Wagar and Meng9,Reference Marks and Bostwick25,Reference Merola, Tomao and Antenucci26} PCA3 is excreted in the urine of prostate cancer patients after DRE, it is a non-invasive diagnostic test capable of predicting biopsy, more accurate than PSA, and can potentially reduce the likelihood of false-positive results. According to Bourdoumis et al., ^{Reference Bourdoumis, Papatsoris, Chrisofos, Efstathiou, Skolarikos and Deliveliotis8} the quantification of PCA3 mRNA levels in urine can predict the outcome of prostatic biopsies with a higher specificity rate in comparison to PSA.

Progensa urinary PCA3 assay (Hologic, Marlborough, Massachusetts) was approved by the US Food and Drug Administration (FDA) in 2012 to help determine the need for repeat prostate biopsies in men over 50 years and with a previous negative biopsy. ^{Reference Cao, Lee, Ning, Handy, Wagar and Meng9,Reference Ferro, Bruzzese and Perdonà16,Reference Cui, Cao and Li21} Since then several studies ^{Reference Cao, Lee, Ning, Handy, Wagar and Meng9,Reference Cui, Cao and Li21,Reference Auprich, Chun and Ward24,Reference Ye, He and Wu27,Reference Wang, Chen, Ren, Shi, Zhu and Zhang30,Reference Hu, Yang and Yang31} have demonstrated PCA3 as a useful prognostic marker for prostate cancer, especially when used in combination with other predictive markers. Auprich et al. ^{Reference Auprich, Chun and Ward24} investigated the significance of PCA3 clinical staging for the identification of pathologic favourable and/or unfavourable features in radical prostatectomy specimens using 305 retrospective clinical and pathologic data of radical prostatectomy patients. They reported improvement in the diagnostic accuracy of externally validated nomograms in patients with elevated PSA and undergoing biopsy using urinary PCA3. Furthermore, they observed that PCA3 independently predicted small-volume disease (<0.5 ml) and pathologically confirmed insignificant prostate cancer. They suggested that PCA3 should be considered as a means to improve risk stratification of low-volume disease and insignificant prostate cancer. However, they reported that PCA3 is not associated with locally advanced disease and has limitations in the prediction of pathologically advanced or aggressive disease at radical prostatectomy. Nevertheless they demonstrated that combining PCA3 with other new biomarkers can improve the diagnostic and prognostic accuracy. Wang et al. ^{Reference Wang, Chen, Ren, Shi, Zhu and Zhang30} also evaluated the effectiveness of urinary PCA3 to predict biopsy results in 500 patients undergoing initial prostate biopsy. They reported that PCA3 score is capable of discriminating positive from negative prostate biopsy outcomes but is not correlated with the aggressiveness of the disease. Cao et al. ^{Reference Cao, Lee, Ning, Handy, Wagar and Meng9} retrospectively reviewed data of 271 patients with elevated PSA, strong family history and/or abnormal DRE results to assess the clinical efficacy of urinary PCA3. They reported that PCA3 score is a significant predictor of prostate biopsy outcome, improves the diagnostic sensitivity and specificity of PSA and the combination of PCA3 score with serum PSA results gives better overall performance than PSA alone in identification of prostate cancer in high-risk patients. Using an optimal PCA3 cut-off score of 30, they reported that the PCA3 assay could have prevented 57.4% of unnecessary invasive biopsies in the overall cohort and about 70.3% in a patient subgroup with PSA level of 4–10 ng/mL. Cui et al. ^{Reference Cui, Cao and Li21} in a meta-analysis study evaluated 46 clinical trials comprising 12,295 urine PCA3 test data to assess the clinical value of urine PCA3 test in the diagnosis of prostate cancer. They reported pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio and diagnostic odds ratio of 65%, 73%, 2·23, 0·48 and 5·31 respectively. They concluded that urine PCA3 test has acceptable sensitivity and specificity for the diagnosis of prostate cancer and can be used as a non-invasive diagnostic method. Hu et al. ^{Reference Hu, Yang and Yang31} conducted a systematic review and meta-analysis studies to determine the diagnostic role of urine PCA3 in detecting prostate cancer and reported sensitivity, specificity, positive diagnostic likelihood ratio, negative diagnostic likelihood ratio and pooled diagnostic odds ratio of 0·57, 0·71, 2·12, 0·55 and 3·93 respectively in the diagnosis of prostate cancer using urine PCA3.

4Kscore Test

The 4Kscore (OPKO Lab, Nashville, TN) test is a serum-based test to determine a patient’s risk of having an aggressive (Gleason ≥7) high-grade prostate cancer if a prostate biopsy were to be performed. It uses a combination of four kallikrein-based markers (i.e. total PSA (tPSA), free PSA (fPSA), intact PSA and human kallikrein 2) as well as patient’s clinical information (i.e. age, prior biopsy results and DRE status) in an algorithm to predict prostate cancer biopsy outcome if it were to be performed. ^{Reference Zappala, Dong and Linder6,Reference Alford, Brito, Yadav, Yadav, Tewari and Renzulli13–Reference Punnen, Pavan and Parekh15,Reference Kornberg, Cooperberg, Spratt and Feng22,Reference Konety, Zappala and Parekh32} The human kallikrein 3 (commonly known as PSA) and the human kallikrein 2 (hK2) are both prostate-derived kallikreins which are overexpressed in prostate cancer and are responsible for altering cell growth regulation, cell invasion and angiogenesis and consequently play a role in the disease progression and aggressiveness. ^{Reference Alford, Brito, Yadav, Yadav, Tewari and Renzulli13,Reference Punnen, Pavan and Parekh15} The PSA circulates in the blood in the form of glycoprotein and binds to protease inhibitors (i.e. bound PSA); however, there is a small fraction that does not bind to the proteases [known as the free PSA (fPSA)] and the total PSA therefore comprises of both the free and bound PSA in the blood. According to Kornberg et al., ^{Reference Kornberg, Cooperberg, Spratt and Feng22} immunoassay tests related to these prostate-specific kallikreins have proven useful for the detection of aggressive disease. The 4Kscore test is reported to be suitable for patients who are opting for a preliminary prostate biopsy due to an abnormal DRE result or an elevated PSA level or for patients with prior negative biopsy results and elevated PSA levels. The accuracy of the 4Kscore test in the detection of clinically significant cancer (Gleason score ≥7) and its potential to reduce the number of unnecessary biopsies have been confirmed by several investigators. ^{Reference Alford, Brito, Yadav, Yadav, Tewari and Renzulli13,Reference Punnen, Pavan and Parekh15,Reference Kornberg, Cooperberg, Spratt and Feng22,Reference Konety, Zappala and Parekh32,Reference Zhuang and Johnson33} According to Punnen et al., ^{Reference Punnen, Pavan and Parekh15} a high 4Kscore test may be useful for the selection of patients with a high probability of having aggressive disease and who would benefit from a prostate biopsy, whereas patients with low 4Kscore test may safely defer biopsy. The 4Kscore has the potential for a personalised risk assessment of having aggressive prostate cancer and the risk of developing distant metastatic disease. It has also been proven to be a useful complementary tool for the shared decision making between patients and their physicians regarding prostate biopsy and treatment. Vickers et al. ^{Reference Vickers, Cronin and Roobol34} caution that frozen or rethawed serum samples of kallikreins will tend to degrade the kallikrein structure and therefore could affect or compromise the potential accuracy of the 4Kscore panel in detecting aggressive prostate cancer if frozen and rethawed serum samples of kallikreins are used instead of fresh samples.

Several studies ^{Reference Zappala, Dong and Linder6,Reference Konety, Zappala and Parekh32,Reference Vickers, Cronin and Roobol34,Reference Vickers, Cronin and Aus35} have assessed the clinical utility of the 4Kscore test for accurate prediction of prostate biopsy outcome in patients with elevated PSA levels and also to determine the risk of high-grade, aggressive (Gleason ≥7) disease. Konety et al. ^{Reference Konety, Zappala and Parekh32} conducted a clinical efficacy study to assess the influence of 4Kscore test on the decision to perform prostate biopsy in 611 patients who were referred to urologists due to abnormal PSA level and/or DRE results. They reported that the 4Kscore test results greatly influenced biopsy decisions in 88·7% of patients, reduced overall prostate biopsy rate by 65% and by 94% in patients with low-risk 4Kscore results. They observed that a higher 4Kscore test was associated with greater likelihood of a patient having a prostate biopsy and the risk category was strongly associated with biopsy pathology. Vickers et al. ^{Reference Vickers, Cronin and Aus35} investigated the potential of a panel of the four kallikrein markers in blood to predict prostate biopsy outcome in previously unscreened 740 patients with elevated total PSA (≥3 ng/mL) undergoing biopsy. They reported that the addition of 4Kscore to a model that included age and PSA improved the prediction of prostate cancer at biopsy and a further improvement of the prediction for a model with age, PSA and DRE. When they set a 20% risk of having prostate cancer as the threshold for biopsy, they estimated a 57% reduction in the number of biopsies. They concluded that the 4Kscore can predict biopsy outcome in patients with elevated PSA, and this multivariable model has the potential to identify patients who should be advised to undergo biopsy and patients who might be advised to continue screening, but defer biopsy until there is stronger evidence of malignancy. In a more recent study, Vickers et al. ^{Reference Vickers, Cronin and Roobol34} reproduced their previous findings in a larger, independent, population-based cohort of 2,914 previously unscreened patients undergoing biopsy as a result of elevated PSA (≥3 ng/mL) and further emphasised that application of the 4Kscore panel would significantly reduce biopsy rates. Zappala et al. ^{Reference Zappala, Dong and Linder6} performed a study to assess the clinical validity of 4Kscore test with or without DRE information in 1,012 patients scheduled for prostate biopsy and reported that the predictive accuracy of 4Kscore with DRE was similar to 4Kscore without DRE. They concluded that patients who are suspected of aggressive prostate cancer can safely re-define their risk with the 4Kscore test even in the absence of a DRE result. According to Punnen et al., ^{Reference Punnen, Pavan and Parekh15} using the 4Kscore test to select patients with a significant probability of having aggressive prostate cancer would potentially avoid premature death from the disease, while at the same time safely preclude patients who do not require unnecessary invasive intervention.

Prostate Health Index (PHI)

The Beckman Coulter Prostate Health Index (PHI) test, approved by the US Food and Drug Administration in June 2012, is a diagnostic blood test that combines three quantitative kallikrein immunoassays; tPSA, fPSA and the (−2) pro-PSA isoform (p2PSA) into a single numerical score as shown below.

$$PHI\;Score = \;{{p2PSA} \over {fPSA}}\sqrt {tPSA} $$

The PHI score is intended to assist in clinical decision making regarding prostate biopsy in order to help reduce the number of unnecessary biopsies in patients with elevated PSA. ^{Reference White, Shenoy and Tutrone3,Reference Lazzeri, Haese and Abrate5,Reference Hsieh, Chang and Yang18–Reference Loeb and Catalona20,Reference Loeb, Shin and Broyles36} According to Loeb et al., ^{Reference Loeb, Shin and Broyles36} the PHI test results are currently reported in four distinct ‘risk bin’ categories: 0–24·9; 25·0–34·9; 35·0–54·9 and >5·0 with the corresponding probability of prostate cancer detection on biopsy of 11·0, 18·1, 32·7 and 52·1%, respectively. Friedersdorff et al. ^{Reference Friedersdorff, Groß and Maxeiner37} reported an association between PHI and prostate cancers with unfavourable prognostic characteristics and at 90% sensitivity the specificity of PHI to detect prostate cancer is about 31·6%. Several studies have shown that PHI is a promising new prostate marker, it is cost-effective, preferentially detects aggressive carcinomas, it is consistently associated with Gleason score and has high specificity thereby can reduce a number of unnecessarily performed biopsies. ^{Reference Lazzeri, Haese and Abrate5,Reference Bourdoumis, Papatsoris, Chrisofos, Efstathiou, Skolarikos and Deliveliotis8,Reference Ferro, Bruzzese and Perdonà16,Reference Fossati, Lazzeri and Haese17,Reference Lepor, Catalona and Loeb19,Reference Loeb and Catalona20,Reference Friedersdorff, Groß and Maxeiner37,Reference Abrate, Lazzeri and Lughezzani38} Other studies have also reported that PHI outperforms both tPSA and %fPSA (i.e. fPSA as percentage of tPSA) for prostate cancer detection, has improved prediction of clinically significant disease outcome on biopsy in patients with PSA 4–10 ng/mL and negative DRE and can predict biopsy outcome reclassification in patients enrolled in active surveillance. ^{Reference Lazzeri, Haese and Abrate5,Reference Hsieh, Chang and Yang18–Reference Loeb and Catalona20,Reference Loeb, Shin and Broyles36} According to Fossati et al. ^{Reference Fossati, Lazzeri and Haese17} and Lepor et al., ^{Reference Lepor, Catalona and Loeb19} %p2PSA (i.e. p2PSA as percentage of free PSA) and PHI are more accurate in determining the presence and aggressiveness of prostate cancer at biopsy compared to tPSA, fPSA and %fPSA. Furthermore, Lazzeri et al. ^{Reference Lazzeri, Haese and Abrate5} also reported that p2PSA, %p2PSA and PHI will potentially improve both the discrimination between patients with and without prostate cancer and between cases of aggressive and non-aggressive disease. Friedersdorff et al. ^{Reference Friedersdorff, Groß and Maxeiner37} reported that using PHI score in prostate cancer screening can reduce biopsies by 15–41% when compared to screening with only tPSA and Loeb et al. ^{Reference Loeb, Shin and Broyles36} also indicated that deferring prostate biopsy for patients with PHI <24 would preclude 36–41% of unnecessary biopsies. According to Lepor et al., ^{Reference Lepor, Catalona and Loeb19} the addition of PHI to a multivariable model with age, prostate volume, DRE and prostate biopsy history resulted in a statistically significant 7% gain in predictive accuracy on biopsy and should be considered as part of the standard urologic tool for biopsy decisions, risk stratification and treatment selection. Moreover, Loeb and Catalona ^{Reference Loeb and Catalona20} also indicated that PHI has the ability to predict the likelihood of disease progression during active surveillance, thereby providing an additional non-invasive modality to potentially select and monitor this patient population.

Several studies have examined the relationship of PHI with clinically-significant disease on biopsy and with adverse prostatectomy outcomes, evaluated the specificity of PHI for clinically-significant prostate cancer in biopsy-naive patients and also demonstrated a significant relationship between PHI and adverse tumour features such as pathologic stage, grade, tumour volume and composite at radical prostatectomy outcomes of clinically significant disease. ^{Reference White, Shenoy and Tutrone3–Reference Lazzeri, Haese and Abrate5,Reference Ferro, Bruzzese and Perdonà16–Reference Lepor, Catalona and Loeb19,Reference Loeb, Shin and Broyles36–Reference Loeb, Sanda and Broyles39} Abrate et al., ^{Reference Abrate, Lazzeri and Lughezzani38} Fossati et al. ^{Reference Fossati, Lazzeri and Haese17} and Lazzeri et al. ^{Reference Lazzeri, Haese and Abrate5} investigated the sensitivity, specificity and accuracy of p2PSA, %p2PSA and PHI to predict prostate cancer at biopsy. All studies reported that p2PSA, %p2PSA and PHI were significantly elevated in patients with prostate cancer compared to those with a negative biopsy. Abrate et al. ^{Reference Abrate, Lazzeri and Lughezzani38} reported that PHI considerably increased the accuracy of the base multivariable model by 8·8% and at a PHI threshold of 35·7, about 32·4% biopsies could have been avoided. Fossati et al. ^{Reference Fossati, Lazzeri and Haese17} and Lazzeri et al. ^{Reference Lazzeri, Haese and Abrate5} also reported that %p2PSA and PHI are the most accurate predictors of prostate cancer at biopsy, achieved independent predictor status, significantly increased the accuracy of multivariate models including PSA and prostate volume by 6·3 and 7·6% respectively and at a PHI threshold of 25·5 a total of 17·2% biopsies could have been avoided. All the studies concluded that PHI is significantly more accurate than current tests in predicting prostate cancer in patients. Ferro et al. ^{Reference Ferro, Bruzzese and Perdonà16} evaluated the potential of PHI to identify prostate cancer at initial biopsy and its correlation with active surveillance eligibility in 300 patients aged >50 years with tPSA of 2–10 ng/ml. They reported PHI is a strong predictor of prostate cancer in patients with tPSA of 2–10 ng/ml and it is significantly more accurate than the currently employed %fPSA tests. In addition, they indicated that PHI is strongly correlated with biopsy outcomes, suggesting a potential role in patients’ selection for active surveillance. Foley et al. ^{Reference Foley, Gorman and Sharifi4} analysed the clinical value of PHI index score and clinical information to inform the decision for prostate biopsy in 250 patients referred for prostate cancer assessment. They reported that the PHI model showed the best correlation between predicted probabilities and actual outcome, has the potential to improve decision making in a clinical setting, was a more discriminant biomarker than PSA and has superior net benefit. Therefore, they suggested that utilising PHI-based risk stratification in the clinical settings will enable urologists to make a more informed decision for each patient. Friedersdorff et al. ^{Reference Friedersdorff, Groß and Maxeiner37} conducted a study to investigate the correlation between PHI, tumour volume and Gleason score in 196 prostate cancer patients treated with radical prostatectomy and reported that PHI is the most accurate predictor of tumour volumes >0·5cc and it correlates more significantly with tumour volume than Gleason score. Furthermore, utilising PHI improved the prediction of larger tumour volumes and subsequently clinically significant cancer and the integration of PHI test into active surveillance is economically beneficial.

Hsieh et al. ^{Reference Hsieh, Chang and Yang18} studied the diagnostic performance of tPSA, fPSA, %fPSA, p2PSA, %p2PSA and PHI to predict the detection of prostate cancer in patients aged 50–75 years with a serum PSA level of 4–10 ng/mL undergoing transrectal prostate biopsy. They reported sensitivity and specificity of 77·8 and 67·8% respectively at a PHI cut-off value of 29·6 and avoided 57·1% of unnecessary biopsy. They concluded that %p2PSA and PHI have better potential to predict prostate biopsy outcome than PSA and %fPSA at initial biopsy in patients with serum PSA of 4–10 ng/mL. Loeb et al. ^{Reference Loeb, Sanda and Broyles39} investigated if PHI improves the specificity for detecting clinically significant prostate cancer and can help reduce prostate cancer over-diagnosis in a multicentre prospective trial in 658 patients aged ≥50 years with PSA of 4–10 ng/ml and normal DRE. They demonstrated that PHI outperformed its individual components for the identification of clinically significant prostate cancer. At a PHI threshold of 28·6, about 30% of biopsies with indolent or no prostate cancer would have been potentially avoided. In another study, Loeb et al. ^{Reference Loeb, Shin and Broyles36} examined whether the addition of PHI improves the performance of the Prostate Cancer Prevention Trial (PCPT) and European Randomised Study of Screening for Prostate Cancer (ERSPC) risk calculators for aggressive prostate cancer in a large multicentre study involving 728 patients with PSA levels of 2–10 ng/mL and a negative DRE. They reported that the addition of PHI significantly improved the predictive accuracy of the PCPT and ERSPC risk calculators for aggressive disease and concluded that using PHI as part of multivariable risk assessment leads to a significant improvement in the detection of aggressive disease, potentially reducing unnecessary biopsy and over-diagnosis. White et al. ^{Reference White, Shenoy and Tutrone3} conducted an observational study to prospectively examine the impact of PHI testing on biopsy decisions in 506 patients receiving a PHI test and 683 men as control group (without PHI) at four large urology clinics. They recorded physicians’ recommendations before and after receiving the PHI test results. They reported that patients who received a PHI test showed a significant reduction in biopsy procedures performed compared to those who did not receive their test. The PHI score also impacted physicians’ patient management plan in 73% of cases, including biopsy deferrals when the PHI score was low, and decisions to perform biopsies when the PHI score indicated an intermediate or high probability of prostate cancer. They concluded that PHI testing significantly impacted physicians’ biopsy decision for patients with tPSA of 4–10 ng/ml and non-suspicious DRE findings and its utilisation resulted in a significant reduction in biopsy procedures performed compared to patients seen by the same urologists who would have met enrolment eligibility but did not receive a PHI test.

TMPRSS2:ERG Gene Fusion

The Transmembrane Serine Protease 2 (TMPRSS2) gene located on the long arm of chromosome 21 at position 22.3 (21q22.3) is a prostate-specific, androgen-response gene that encodes a protein belonging to the serine protease family. The oestrogen-regulated gene (ERG) located on the long arm of chromosome 21 at position 22.2 (21q22.2) is a protein coding gene and encodes a member of the erythroblast transformation-specific (ETS) family of transcription factors. ^{Reference Hossain and Bostwick40–Reference Nam, Sugar and Wang43} The role of TMPRSS2 protein in prostate carcinogenesis depends on the overexpression of ETS transcription factors such as ERG and ETS-translocation variant 1 (ETV1) through gene fusion, and several studies have reported that the TMPRSS2:ERG is the most frequent gene fusion found in 40–80% of human prostate cancers. ^{Reference Hossain and Bostwick40–Reference Zhou, Gao and Han51} Numerous studies have reported that TMPRSS2 is an androgen-responsive gene and as a result ERG overexpression resulting from TMPRSS2:ERG fusion may contribute to the development of androgen-independence in prostate cancer through the disruption of androgen-receptor signalling (ARS), especially in castrate-resistant prostate cancer, ^{Reference Hossain and Bostwick40,Reference Kulda, Topolcan and Kucera41,Reference Yu, Yu and Mani45,Reference Tomlins, Laxman and Varambally49–Reference Zhou, Gao and Han51} and could be exploited for prostate cancer diagnosis in tissue, urine and blood samples. ^{Reference Hossain and Bostwick40} According to Kulda et al., ^{Reference Kulda, Topolcan and Kucera41} TMPRSS2:ERG gene fusion appears in early high-grade prostatic intraepithelial neoplasia (PIN) and it has been observed that its presence in prostate cancer cells promotes the loss of the Phosphate and Tensin Homologue Deleted on Chromosome 10 (PTEN), a tumour suppressor gene, which subsequently leads to the acceleration of disease progression. Current studies have indicated that the identification of TMPRSS2:ERG fusion in prostate cancer suggests that distinct molecular subtypes may further define the risk for the disease progression. ^{Reference Demichelis, Fall and Perner44,Reference Lee, Lee and Park47,Reference Tomlins, Laxman and Varambally49} It has also been reported that the presence of TMPRSS2:ERG gene fusion in prostate tumours is associated with an aggressive phenotype, moderate to poorly differentiated prostate tumours and disease recurrence, progression and prostate cancer-specific death and has potential as a prognostic indicator for prostate cancer. ^{Reference Lee, Lee and Park47,Reference Tomlins, Laxman and Varambally49,Reference Wang, Wang and Zhang50,Reference FitzGerald, Agalliu and Johnson52}

Several studies have investigated the potential of the TMPRSS2:ERG gene fusion as a prognostic biomarker in prostate cancer and its relationship with various clinicopathologic parameters such as age, serum PSA level, Gleason score, histological features and pathologic stage. ^{Reference Kulda, Topolcan and Kucera41–Reference Wang, Wang and Zhang50,Reference FitzGerald, Agalliu and Johnson52} Demichelis et al. ^{Reference Demichelis, Fall and Perner44} investigated the presence of TMPRSS2:ERG fusion and the risk of metastases or prostate cancer-specific mortality in 111 patients with localised disease. They reported a statistically significant association between TMPRSS2:ERG fusion and disease metastases or prostate cancer-specific mortality and concluded that TMPRSS2:ERG gene fusion potentially contributes to a more aggressive disease phenotype, possibly mediated through ERG overexpression. FitzGerald et al. ^{Reference FitzGerald, Agalliu and Johnson52} investigated whether TMPRSS2:ERG fusion is associated with prostate cancer-specific mortality using tumour samples from 372 patients and long-term surveillance after cancer diagnosis. They reported reduced prostate cancer-specific survival in patients whose tumour had multiple copies of TMPRSS2:ERG gene fusion. Hagglof et al. ^{Reference Hägglöf, Hammarsten and Strömvall46} retrospectively examined a large cohort of diagnosed prostate cancer cases which were managed by active surveillance via ERG immunostaining to investigate if ERG staining is associated with tumour characteristics, long-term outcome and differences in tumour stroma morphology. They reported an association between TMPRSS2:ERG fusion and shorter patient survival, alterations in tumour stroma such as increased Platelet Derived Growth Factor Receptor Beta (PDGFRB), vascular density and hyaluronan and decreased Caveolin-1. They concluded that TMPRSS2:ERG gene fusion is associated with a more aggressive prostate cancer phenotype and is supported by changes in the tumour stroma. Kulda et al. ^{Reference Kulda, Topolcan and Kucera41} evaluated the relationship of TMPRSS2:ERG gene fusion status, tumour PSA, PCA3, miR-23b, miR-26a and miR-221 expression levels in combination with preoperative serum PSA level with the risk of prostate cancer recurrence after radical prostatectomy in 108 patients. They observed significantly shorter time to recurrence in patients with high expression of TMPRSS2:ERG and reported that a combination of high level preoperative serum PSA and overexpression of TMPRSS2:ERG fusion transcript in tumour tissue is a promising marker of the risk of recurrence after radical prostatectomy and could potentially discriminate aggressive and life-threatening tumours. Magi-Galluzzi et al. ^{Reference Magi-Galluzzi, Tsusuki and Elson42} examined the prevalence and type of TMPRSS2:ERG gene fusion in prostate cancer from 42 Caucasian, 64 African-American and 44 Japanese patients who underwent radical prostatectomy. They reported that TMPRSS2:ERG gene fusion correlated with ethnicity, marginally correlated with pathologic stage, but not with other clinicopathologic parameters such as age, preoperative PSA levels and Gleason score.

Circulating Tumour Cells (CTCs)

CTCs are a population of rare tumour cells that have detached from either a primary tumour and/or secondary tumour sites and circulating in the peripheral blood or lymphatic vessels of cancer patients and are broadly considered as the main mechanism for tumour metastasis. ^{Reference Broncy and Paterlini-Bréchot53–Reference Hegemann, Stenzl, Bedke, Chi, Black and Todenhöfer57} According to Galletti et al. ^{Reference Galletti, Portella, Tagawa, Kirby, Giannakakou and Nanus58} and Hegemann et al., ^{Reference Hegemann, Stenzl, Bedke, Chi, Black and Todenhöfer57} CTCs have been detected in the peripheral blood of patients with localised or locally advanced prostate cancer without clinical evidence of metastatic disease, suggesting that these cells may act as the seeds for the development of metastases. Broncy and Paterlini-Brechot ^{Reference Broncy and Paterlini-Bréchot53} reported that the prognostic value of CTCs is well established for metastatic prostate cancer and its detection could provide a valuable complement to PSA or other prostate cancer detection tests to identify patients with more aggressive disease. Danila et al. ^{Reference Danila, Fleisher and Scher56} also reported that post-treatment CTC number in castration-resistant prostate cancer (CRPC) patients is a stronger prognostic factor for survival than a 50% decline in PSA. Other studies have also reported that although CTCs can be found in up to about 75% of patients with advanced prostate cancer, its detection is usually limited in patients with non-metastatic disease. ^{Reference Hegemann, Stenzl, Bedke, Chi, Black and Todenhöfer57,Reference Salami, Singhal and Spratt59} According to Galletti et al. ^{Reference Galletti, Portella, Tagawa, Kirby, Giannakakou and Nanus58} and Hegemann et al., ^{Reference Hegemann, Stenzl, Bedke, Chi, Black and Todenhöfer57} CTC detection in the peripheral blood of patients may imply metastatic disease progression thereby suggesting early treatment before metastatic cells become clinically detectable with standard radiographic imaging. Numerous studies have observed a consistent correlation between high CTC counts and disease progression, overall survival, unfavourable prognosis and worse clinical outcome and are superior to PSA tests in predicting overall survival. ^{Reference Hegemann, Stenzl, Bedke, Chi, Black and Todenhöfer57–Reference Lorente, Olmos and Mateo60} According to Galletti et al., ^{Reference Galletti, Portella, Tagawa, Kirby, Giannakakou and Nanus58} recent developments in CTC detection and downstream molecular characterisation would potentially provide a more sensitive and reliable approach for the detection of early-stage prostate cancers and discriminate indolent from aggressive disease and could potentially be integrated into prostate cancers screening programmes. Although, CTCs are currently considered a very promising biomarker to aid cancer diagnosis, treatment decision and patient follow-up, its clinical utility is limited by its extreme low abundance in the peripheral blood (estimated to be 1 cell in 106–109 nucleated cells) therefore making its detection and characterisation technically challenging. ^{Reference Chen, Lu, Cheng, Tseng, Figlin and Posadas54–Reference Galletti, Portella, Tagawa, Kirby, Giannakakou and Nanus58}

The clinical utility of monitoring CTC changes with treatment as a surrogate biomarker for overall survival; determining prognosis; monitoring or predicting treatment efficacy and clinical outcome and as a source of tissue to identify predictive markers of drug sensitivity to guide treatment selection have been investigated. ^{Reference Choi, Lim and Kyung55,Reference Salami, Singhal and Spratt59,Reference Lorente, Olmos and Mateo60} Choi et al. ^{Reference Choi, Lim and Kyung55} evaluated the clinical effectiveness of CTC tests in 17 healthy volunteers and 45 patients with localised prostate cancer undergoing treatment or under active surveillance. They reported that CTC positivity in patients was associated with Gleason pattern 5, therefore suggesting that active treatment should be considered for low-risk disease patients undergoing active surveillance but have positive expression of epithelial cell adhesion molecule (EpCAM+) CTCs because of a risk of up-staging and up-grading. Lorente et al. ^{Reference Lorente, Olmos and Mateo60} conducted a study to determine the significance of elevated CTCs as an indicator of disease progression in 511 patients with pre-treatment CTCs <5. They reported that increased CTC count in the first 12 weeks of treatment was independently associated with worse overall survival in patients with baseline CTCs <5 treated with abiraterone or chemotherapy. Therefore CTCs can potentially identify patients with low baseline CTC counts (<5) who will not benefit from treatment. Salami et al. ^{Reference Salami, Singhal and Spratt59} evaluated the potential of CTCs to predict treatment response for high-risk localised prostate cancer in 49 patients and reported that CTC counts can be identified in patients with high-risk localised prostate cancer before definitive therapy. They concluded that phenotypic and genomic characterisation of pre-treatment CTCs could provide an added means of risk stratification of patients with newly diagnosed high-risk disease and potentially help identify patients who could require multimodal therapy.

Phosphate and Tensin Homologue Deleted on Chromosome 10 (PTEN)

PTEN is a tumour suppressor gene located on the long arm of chromosome 10 at position 23.31 (10q23.31) and encodes a phosphatidylinositol-3,4,5-trisphosphate 3-phosphatase protein with lipid phosphatase activity. ^{Reference Pourmand, Ziaee and Abedi61,Reference Ye, Zhao and Kang62} Numerous studies have reported that PTEN facilitates functions related to smooth muscle cell differentiation, angiogenesis and T-reg cell stability in normal physiological conditions. ^{Reference Pourmand, Ziaee and Abedi61–Reference Stefano and Giovanni65} According to Stefano and Giovanni, ^{Reference Stefano and Giovanni65} PTEN frequently loses its function through deletion, mutation, transcriptional silencing or protein instability and the loss of function has consequent alterations in key pathways implicated in cell proliferation, migration, survival and genomic stability. PTEN has been documented by several studies to play an essential role in tumourigenesis and both its inactivation and mutation are reported to influence tumour cell proliferation, apoptosis and cell cycle progression. ^{Reference Pourmand, Ziaee and Abedi61–Reference Jamaspishvili, Berman and Ross64} Jamaspishvili et al. ^{Reference Jamaspishvili, Berman and Ross64} and Tosoian et al. ^{Reference Tosoian, Guedes and Morais66} reported that the inactivation of PTEN by deletion or mutation is one of the most common aberrations in prostate cancer, occurring in approximately 20–50% of primary tumours at radical prostatectomy and in about 50% of castration-resistant tumours. The clinical significance of PTEN loss or inactivation has been associated with adverse pathology, shorter time to biochemical recurrence and increased frequency of metastatic disease. In addition, PTEN loss or inactivation also promotes castration-resistant tumour growth through upregulation of oncogenic Akt/mTOR signalling suppression of AR transcription factor activity and inhibition of AR-regulated negative feedback of Akt. ^{Reference Pourmand, Ziaee and Abedi61–Reference Jamaspishvili, Berman and Ross64,Reference Tosoian, Guedes and Morais66,Reference Lotan, Heumann and Rico67} According to Jamaspishvili et al. ^{Reference Jamaspishvili, Berman and Ross64} and Pourmand et al., ^{Reference Pourmand, Ziaee and Abedi61} PTEN has the potential as a useful genomic marker to distinguish indolent from aggressive prostate cancer in patients with clinically localised tumours.

Ahearn et al. ^{Reference Ahearn, Pettersson and Ebot63} conducted a broad patho-epidemiology study in 1,044 prostate cancer patients to investigate the association of PTEN loss [assessed by immunohistochemistry (IHC) assay) with fatal disease progression and the potential for TMPRSS2:ERG gene fusion to modify the role of PTEN loss in fatal disease progression. They reported that patients whose tumours showed complete PTEN loss had a statistically significant elevated risk for fatal disease progression, especially in patients with ERG fusion-negative in comparison with patients whose tumours showed PTEN intact or heterogeneous loss or ERG fusion-positive. Furthermore, they observed that complete PTEN loss was associated with elevated risk for biochemical recurrence and may be useful for prostate cancer risk stratification. Pourmand et al. ^{Reference Pourmand, Ziaee and Abedi61} analysed 51 formalin-fixed paraffin-embedded prostate cancer specimens for PTEN mutations to investigate the role of PTEN gene in disease progression. They observed that patients with positive PTEN mutation had a significantly greater Gleason score, lower survival rate, higher tendency to metastasis and higher PSA. Ye et al. ^{Reference Ye, Zhao and Kang62} retrospectively investigated the expression levels and significance of PTEN in the blood of 84 prostate cancer patients and 84 healthy volunteers. They reported that PTEN expression levels are significantly related with distant metastasis of disease, preoperative PSA level, tumour size and pathological stage and can be used as a biological indicator for disease diagnosis, severity analysis and efficacy evaluation. Tosoian et al. ^{Reference Tosoian, Guedes and Morais66} performed a case-control study to investigate the association between PTEN status on biopsy and subsequent clinicopathologic outcome in 223 prostate cancer patients with Gleason score of 6 who were initially managed with active surveillance. They observed that PTEN loss on diagnostic biopsy was common among patients who had grade reclassification during active surveillance and could aid to identify patients with aggressive disease. They concluded that PTEN IHC testing is simple, inexpensive and holds promise as a tool to provide valuable information for clinicians and patients. Lotan et al. ^{Reference Lotan, Heumann and Rico67} studied PTEN IHC and PTEN fluorescence in situ hybridisation (FISH) in 9,033 tumours in radical prostatectomy patients with clinical follow-up data and reported similarities between both IHC and FISH results. They concluded that PTEN IHC and PTEN FISH could potentially be used as complementary screening tools for PTEN loss in prostate cancer.

ConfirmMDx (DNA methylation of GSTP1, APC and RASSF1 genes)

ConfirmMDx for prostate cancer (MDxHealth, Inc, Irvine, CA) is a molecular test that has clinically been validated for the detection of prostate cancer in tissues from prostate cancer-negative biopsies. It is an epigenetic multiplex tissue assay that uses methylation-specific polymerase chain reaction (PCR) to measure the epigenetic status of three prostate cancer-associated gene biomarkers in residual histopathologically prostate cancer negative biopsy core tissue samples to determine if a repeat biopsy can safely be avoided. ^{Reference Van Neste, Bigley and Toll68–Reference Wojno, Costa and Cornell72} The three gene biomarkers are the Glutathione S-Transferase Pi 1 (GSTP1) located on the long arm of chromosome 11 at position 13.2 (11q13.2), Adenomatous Polyposis Coli (APC) located on the long arm of chromosome 5 at position 22.2 (5q22.2) and Ras Association Domain Family Member 1 (RASSF1) which is located on the short arm of chromosome 3 at position 21.31 (3p21.31). The GSTP1 gene is responsible for DNA detoxification and it is reported to be very common in prostate cancers (90% of cases), the APC gene is involved in apoptosis, cell migration and cell adhesion and the RASSF1 gene is involved in cell cycle regulation. Furthermore, these genes are reported to increase the sensitivity of prostate cancer detection, are implicated in disease oncogenesis and are predictive of disease aggressiveness and prognosis. ^{Reference Stewart, Van Neste and Delvenne69–Reference Van Neste, Partin, Stewart, Epstein, Harrison and Van Criekinge71,Reference Mahon, Qu and Lin73} According to Wojno et al., ^{Reference Wojno, Costa and Cornell72} these markers can be used to detect epigenetic abnormalities in tissues adjacent to tumours and thereby can potentially detect occult disease missed by histopathology. The ConfirmMDx test identifies the hypermethylation pattern of the CpG island promoter regions of GSTP1, APC and RASSF1 genes and capitalises on the field effect concept that epigenetic alterations can be observed in histopathologically benign-appearing biopsy core tissues adjacent to cancerous tissue. Wojno et al. ^{Reference Wojno, Costa and Cornell72} reported that this field effect is a strong independent predictor to diagnose prostate cancer. According to Stewart et al. ^{Reference Stewart, Van Neste and Delvenne69} and Partin et al., ^{Reference Partin, Van Neste and Klein70} a methylation negative ConfirmMDx result has about 88–90% negative predictive value for all prostate cancers. ConfirmMDx has been used in the clinical settings to make clinical decisions and the results from this epigenetic assay can potentially guide urologists on decisions regarding the need for repeat biopsy for patients with a previous negative biopsy but who are still considered to be at-risk for prostate cancer.

Stewart et al. ^{Reference Stewart, Van Neste and Delvenne69} used the ConfirmMDx assay to retrospectively investigate the detection of prostate cancer in initial histopathologically negative biopsy samples from 498 patients. These patients were subsequently diagnosed with the disease during a follow-up biopsy in The Methylation Analysis to Locate Occult Cancer (MATLOC) study of which 18% were positive cases and 82% were negative controls. They reported a sensitivity of 68%, specificity of 64 and 90% negative predictive value for ConfirmMDx and stated that the assay performed better clinically compared to initial biopsy histopathology, PSA or DRE alone. In a multivariate model that accounted for patient age, PSA, DRE and first biopsy histopathological characteristics, they reported ConfirmMDx was a significant independent predictor of patient outcome. They concluded that the clinical use of the assay could potentially improve disease diagnosis and reduce the number of unnecessary repeat biopsies. Partin et al. ^{Reference Partin, Van Neste and Klein70} evaluated archived, cancer negative prostate biopsy core tissue samples of 350 patients recruited in the Detection Of Cancer Using Methylated Events in Negative Tissue (DOCUMENT) multicentre trial from five urological centres. All patients had repeated biopsy within 24 months with a negative (controls) or positive (cases) histopathological result and they performed a centralised blinded pathology evaluation of the two biopsy series in all available subjects using ConfirmMDx. They reported a negative predictive value of 88% for ConfirmMDx and in multivariate models that adjusted for age, PSA, DRE, first biopsy histopathological characteristics and race, they demonstrated that ConfirmMDx test is the most significant independent predictor of patient outcome. They concluded that ConfirmMDx is a significant, independent predictor of prostate cancer detection in a repeat biopsy after an initial negative result. Van Neste et al. ^{Reference Van Neste, Partin, Stewart, Epstein, Harrison and Van Criekinge71} evaluated the performance of ConfirmMDx to predict patients at-risk of harboring high-grade disease using 7,899 prostate core biopsies from 803 patients. They reported a 96% negative predictive value for low ConfirmMDx levels in disease-negative biopsies and that ConfirmMDx is a significant and important predictor of high-grade cancer. Furthermore, they developed a risk score which combined ConfirmMDx intensity and traditional clinical risk factors and reported that it improved the identification of patients with high-grade disease, maximally avoided unnecessary repeat biopsies and has strong clinical utility as a repeat biopsy decision making tool. Yonover et al. ^{Reference Yonover, Steyaert and Cohen74} studied the impact of ConfirmMDx methylation-specific PCR test on the management of 605 patients with a prior prostate cancer negative biopsy who were counseled to undergo repeat biopsy due to persistent elevated risk of the disease. They reported repeat biopsy rates for ConfirmMDx positive (297 men) of 32.3% (96/297) and for ConfirmMDx negative (308 men) of 5.8% (15/308). They concluded that ConfirmMDx has a clinical utility since it significantly impacted repeat prostate biopsy decision making in the urology setting.

AR Splice Variants (AR-V7)

AR splice variant 7 (AR-V7) is an alternatively spliced isoform of the AR gene located on the long arm of chromosome X at position 12 (Xq12). The gene contains the DNA-binding domain; however, it lacks the regulatory ligand-binding domain, resulting in constitutive and androgen-independent activation of oncogenic signalling and cell proliferation. ^{Reference Wadosky and Koochekpour75–Reference Xu and Qiu80} Xu and Qiu ^{Reference Xu and Qiu80} reported the potential major role AR-V7 plays in prostate cancer progression and therapeutic resistance and indicated that AR-Vs have elevated expression levels in more aggressive and metastatic tumours and could serve as both a prognostic and predictive marker for metastatic prostate cancer patients. Furthermore, they reported that the expression of AR-V7 in CTC could be used as a predictive marker for therapeutic response to guide treatment selection. Numerous investigators have reported that AR-V7 detection in CTC is associated with resistance to ARS inhibitors such as abiraterone acetate and enzalutamide, but not with taxane chemotherapies such as docetaxel and cabazitaxel. ^{Reference Armstrong, Halabi and Luo77,Reference Scher, Graf and Schreiber78,Reference Xu and Qiu80–Reference Antonarakis, Lu and Luber82} Antonarakis et al. ^{Reference Antonarakis, Lu and Wang81} demonstrated that the detection of AR-V7 mRNA in CTCs from patients with advanced prostate cancer is associated with resistance to enzalutamide and abiraterone. According to Markowski et al., ^{Reference Markowski, Silberstein, Eshleman, Eisenberger, Luo and Antonarakis79} the detection of AR-V7 in CTCs of metastatic castration-resistant prostate cancer (mCRPC) patients is predictive of resistance to androgen-receptor targeted therapies (ATTs). Moreover, AR-V7 positive patients have a significantly shorter overall survival, suggesting a prognostic value of AR-V7 in addition to its use as a predictive biomarker. ^{Reference Markowski, Silberstein, Eshleman, Eisenberger, Luo and Antonarakis79} Antonarakis et al. ^{Reference Antonarakis, Lu and Luber82} reported that the detection of AR-V7 mRNA in CTCs correlated with poor outcomes from the use of abiraterone and enzalutamide in patients with CRPC.

Markowski et al. ^{Reference Markowski, Silberstein, Eshleman, Eisenberger, Luo and Antonarakis79} conducted a retrospective study on the clinical utility of the Clinical Laboratory Improvement Amendments (CLIA)-Grade AR-V7 testing in patients with mCRPC. They used a questionnaire-based survey to investigate how AR-V7 testing results are used to influence clinical practice. They reported an increase in AR-V7 detection in patients with prior exposure to ATT. Furthermore, the overall patient management was influenced by AR-V7 testing in about 53% of patients with AR-V7 positivity, and the AR-V7-positive patients were commonly switched from ATT to taxane chemotherapy. Moreover, patients who had a change in management due to AR-V7 testing were significantly more likely to achieve a physician-reported 50% decline in PSA response on next-line therapy. They concluded that AR-V7 testing influences physician’s clinical decision making and AR-V7-positive patients were preferentially managed with taxane therapy which improved patient’s outcome. Scher et al. ^{Reference Scher, Graf and Schreiber78} conducted a blinded correlative study on the clinical utility of the nuclear-localised AR-V7 protein in CTCs as a marker to determine differential overall survival among 142 patients with histologically confirmed mCRPC treated with taxanes versus ARS inhibitors. They observed among the high-risk patients that the AR-V7-positive patients treated with taxanes had significantly longer overall survival relative to those treated with ARS inhibitors. Furthermore, AR-V7-negative patients treated with ARS inhibitors had superior overall survival relative to those treated with taxanes. They concluded that nuclear-localised AR-V7 protein in CTCs can identify patients who may live longer with taxane chemotherapy versus ARS inhibitor treatment. Antonarakis et al. ^{Reference Antonarakis, Lu and Luber82} used a CTC-based AR-V7 mRNA assay to investigate AR-V7 status in 202 patients receiving abiraterone or enzalutamide. They also evaluated the clinical significance of AR-V7 in both the first-line and second-line novel hormonal therapy (NHT) and the prognostic significance in these settings. They reported that CTC and AR-V7-positive patients are more likely to have Gleason scores ≥8, metastatic disease at diagnosis, higher PSA, prior abiraterone or enzalutamide use, prior taxane use and Eastern Cooperative Oncology Group (ECOG) score ≥1. Furthermore, they observed that clinical outcomes were best for CTC-negative patients, intermediate for CTC-positive and AR-V7-negative patients, and worse for CTC-positive and AR-V7-positive patients. They concluded that CTC-based AR-V7 mRNA detection can potentially predict outcomes in patients with CRPC receiving first-line and second-line NHT. Armstrong et al. ^{Reference Armstrong, Halabi and Luo77} evaluated a multicentre, prospective-blinded study of 118 high-risk mCRPC patients starting abiraterone acetate or enzalutamide treatment to validate the prognostic significance of baseline CTC AR-V7 on the basis of radiographic or clinical progression free-survival using the Johns Hopkins University modified-Adna Test CTC AR-V7 mRNA assay and the Epic Sciences CTC nuclear-specific AR-V7 protein assay. They reported that AR-V7 detection by both the Johns Hopkins and Epic AR-V7 assays was independently associated with shorter progression free-survival and overall survivor. They concluded that detection of AR-V7 in CTCs by two blood-based assays is independently associated with shorter progression free-survival and overall survivor with abiraterone or enzalutamide, and patients with mCRPC should be offered alternative treatments.

ExoDx Prostate IntelliScore (EPI)

The ExoDx^TM Prostate IntelliScore (EPI, Exosome Diagnostics, Waltham, MA, USA) test is a non-invasive, non-DRE first-catch urine-based risk assessment tool intended for men aged 50 years or older who have PSA levels 2–10 ng/mL and are being considered for initial biopsy. ^{Reference Tutrone, Donovan and Torkler83–Reference Donovan, Noerholm and Bentink86} The EPI test utilises exosomal RNA expression levels of three biomarkers, namely PCA3, TMPRSS2:ERG and SAM pointed domain containing ETS transcription factor (SPDEF) and an algorithm independent of clinical features to provide a risk score that discriminates high-grade disease (Gleason score ≥7) from low-grade (Gleason score ≤6) and benign disease. ^{Reference Tutrone, Donovan and Torkler83–Reference Press, Schulster and Bjurlin87} PCA3 and ERG are overexpressed in prostate cancer and have been found in post-DRE urine exosomes, whereas SPDEF is expressed in normal prostate epithelium and is considered a prostate cancer tumour suppressor gene. ^{Reference Cui, Cao and Li21,Reference Adamo and Ladomery88–Reference Nilsson, Skog and Nordstrand90} The EPI test has the potential to help reassure patients to avoid a prostate biopsy or improve patients’ compliance with a physician’s recommendation. McKiernan et al. ^{Reference McKiernan, Donovan and Margolis84} demonstrated that the addition of this urine exosomal assay to patients’ standard of care variables such as PSA level, age, race and family history of prostate cancer is associated with improved discriminative ability between low- and high-grade disease. Furthermore, a risk score above the validated cut-point of 15·6 is associated with an increased likelihood of Gleason score ≥7 on a biopsy whereas a score below the cut-point is associated with a decreased likelihood of Gleason score ≥7. McKiernan et al. ^{Reference McKiernan, Donovan and Margolis84} and Donovan et al. ^{Reference Donovan, Noerholm and Bentink86} reported that the validated cut-off of 15·6 yields a negative predictive value greater than 90% and a sensitivity of 92%. Tutrone et al. ^{Reference Tutrone, Donovan and Torkler83} reported that the EPI scores are correlated to high-grade prostate cancer and the validated 15·6 cut-point is designed to reduce unnecessary biopsies.

Donovan et al. ^{Reference Donovan, Noerholm and Bentink86} investigated the efficacy of the EPI assay to discriminate high-grade prostate cancer from low-grade or benign disease using first-catch urine samples from 195 patients aged ≥40 years scheduled for an initial or repeat prostate needle biopsy due to suspicious DRE and/or PSA levels. They reported EPI score had 97·5% negative predictive value and 95·2% sensitivity for high-grade disease. Furthermore, the EPI score alone had significant accuracy in predicting biopsy results for high-grade disease independent of standard of care parameters such as PSA level, age, race and family history of biopsy. They concluded that EPI is a novel non-invasive, molecular assay for predicting high-grade prostate cancer in patients with serum PSA levels in the ‘gray zone’ presenting for an initial biopsy. In addition they reported that the utilisation of EPI in the biopsy decision process could potentially reduce the number of prostate biopsies for clinically insignificant disease. McKiernan et al. ^{Reference McKiernan, Donovan and O’Neill85} investigated the prognostic value of the EPI score plus standard of care to standard of care alone for the discrimination of high-grade prostate cancer in 774 patients aged ≥50 years, scheduled for an initial or repeated prostate needle biopsy due to suspicious DRE findings and/or PSA levels of 2.0–20.0 ng/mL. They reported that in both the training (255 patients) and independent validation (519 patients) cohorts the EPI score plus standard of care was associated with improved discrimination of high-grade prostate cancer compared to standard of care alone. In the training cohort, they reported a negative predictive value of 96% for high-grade disease using an EPI cut-off value of 15·6 and that 20% of prostate biopsies could have been avoided. However, in the validation cohort with the same cut-off value, they reported a negative predictive value of 91% and sensitivity of 92% and that 27% of biopsies could have been avoided. They concluded that EPI score is associated with improved identification of patients with higher-grade prostate cancer among patients with elevated PSA levels and could reduce the total number of unnecessary biopsies. In a two-phase adaptive clinical utility study comparing EPI results with biopsy outcomes in 503 patients age 50 years or older and PSA of 2–10 ng/ml scheduled for initial prostate biopsy, McKiernan et al. ^{Reference McKiernan, Donovan and Margolis84} assessed the performance and utility of the EPI assay against standard clinical parameters for discriminating high-grade prostate cancer from low-grade and benign disease on initial biopsy. They reported that EPI is superior to an optimised model of standard clinical parameters for predicting high-grade disease. Furthermore, they reported that using a validated EPI cut-point of 15·6 had a negative predictive value of 89% and a sensitivity of 93%, and avoided 26% of unnecessary prostate biopsies. When they used a higher cut-off of 20, they reported a negative predictive value of 89% and a sensitivity of 89% and avoided 40% of unnecessary prostate biopsies. Tutrone et al. ^{Reference Tutrone, Donovan and Torkler83} assessed the clinical utility of the EPI test in a prospective, randomised, blinded, two-armed study that enrolled 1,094 patients considered for prostate biopsy based on standard clinical criteria. All patients had an EPI test; however, they were randomised into EPI and control arms, where only the EPI arm received the EPI result as part of the decision for biopsy. They reported that the biopsy decisions of 68% of urologists were influenced by EPI results and the EPI arm detected 30% more high-grade prostate cancer than the control arm. They concluded that the EPI test can easily be incorporated into clinical practice and would potentially improve the identification of clinically significant disease and reduce biopsies.