Active Surveillance in Prostate Cancer: Role of Available Biomarkers in Daily Practice

Prostate cancer (PCa) is the most commonly diagnosed cancer in men. The diagnosis is currently based on PSA levels, which are associated with overdiagnosis and overtreatment. Moreover, most PCas are localized tumours; hence, many patients with low-/very low-risk PCa could benefit from active surveillance (AS) programs instead of more aggressive, active treatments. Heterogeneity within inclusion criteria and follow-up strategies are the main controversial issues that AS presently faces. Many biomarkers are currently under investigation in this setting; however, none has yet demonstrated enough diagnostic ability as an independent predictor of pathological or clinical progression. This work aims to review the currently available literature on tissue, blood and urine biomarkers validated in clinical practice for the management of AS patients.


Introduction
In 2020, 19.3 million new cases of cancer were diagnosed worldwide, according to the International Agency for Research of Cancer, of which 1,414,259 were of prostate cancer (PCa). These figures place PCa in the fourth position in incidence when combining both sexes, preceded by breast, lung and colorectum cancer [1]. When only men are considered, PCa represents the second most frequent malignancy after lung cancer (14.1% and 14.3%, respectively), being considered a global health problem, due to both the volume of population affected and its economic impact on Health Systems.
Almost 90% of new PCa are localized and they are clinically classified in three risk groups, based on the serum prostate-specific antigen (PSA) levels, local clinical stage and histological aggressiveness provided by the Gleason Score (GS): high risk, intermediate risk, low risk and very low risk [2]. All clinical guidelines, such as European Society of Medical Oncology (ESMO) [3], National Comprehensive Cancer Network (NCCN) [4] and 2020 EAU-EANM-ESTRO-ESUR-SIOG [5], follow this classification with very little variation.

Evidence Acquisition
A comprehensive literature search from January 2013 to March 2021 was performed in PubMed, including articles written in English language, reporting on PCa diagnosis and follow-up biomarkers. A specific search strategy was designed combining the following keywords: "prostate cancer", "active surveillance", "tissue biomarkers", "blood biomarkers", "urine biomarkers" and "clinical significance". In particular, the following search blocks were used for the PubMed database (active surveillance) AND (prostate cancer) AND (biomarker) AND [(tissue) OR (blood) OR (urine)] AND (clinical significance). Due to the exigent search criteria used, mainly based on the addition of the "clinical significance" label, secondary sources were also examined in a descriptive manner. Cross-referenced potentiality relevant articles, not identified in the primary search, were also considered and hand-picked. Case reports, editorials, letters, congress abstract and congress communications were not eligible.
Original research articles were curated based on favouring large sample sizes, independent validation and patients directly included in AS programs and not just as per its role in optimizing PCa diagnosis. When biomarkers were studied in diagnosis, we focused our interest in those cohorts already in AS and its role in its clinical management. After exclusion of duplicates and articles unrelated to the topic of this review, 267 fulltext records were screened, and finally, 26 papers accomplished final eligibility. Then, 39 other papers were added that focus on commercially available markers related to the topic. The article selection process is shown in a workflow diagram (Figure 1).

PSA
Nobody questions the role of PSA in PCa screening, with the latest publication of the European Randomized Screening in Prostate Cancer setting its irreplaceable importance [19]. However, once diagnosed, in those PCa candidates for AS, its role as biomarker has been questioned due to its variability and has been preferably studied related to its changes with time, prostate volume or other isoforms.

PSA kinetics
Diagnostic and follow-up of PCa patients based on PSA measurements are still the most common strategies, although PSA has been demonstrated as an unspecific biomarker in the AS setting. The rate of PSA change over time, known as PSA dynamics or kinetics (PSAk), was conceptually introduced by Carter in 1992 [20]. PSAk has been shown to overcome PSA limitations and to predict PCa reclassification in men enrolled in AS programs.
Cooperberg et al., calculated PSAk using a linear mixed-effect model, in which the natural logarithm of PSA (ln[PSA]) was modelled as a linear function of time since the diagnosis, with a random intercept indicating the individual-specific ln[PSA] at diagnosis and a random slope reflecting the individual-specific rate of change over time. It was tested in a multicentre cohort with long-term follow-up, suggesting that collecting PSA measurements over time could be clinically useful at predicting outcomes in men with PCa on AS [21].
Related to PSAk, PSA doubling-time (PSADT) is the number of years over which a certain level of PSA increases by a factor of two and is calculated as DT = ln(2)/m, where m is the slope of the regression of ln[PSA] over time [22]. In the prospective multicentre Canary Prostate Active Surveillance Study (PASS), PSADT < 36 months was originally a criterion for progression, and some authors recommended this parameter for the detection of aggressive tumours during AS [23]. However, it was found to be unspecific [24] and must be evaluated as a part of the PSAk linear model.
On the other hand, PSA velocity (PSAV) represents a change in PSA level over time and is calculated by linear regression of untransformed PSA values. By calculating the number of times that serial PSAV measurements pass a threshold, Patel et al. developed the PSAV Risk Count (PSAV RC) score. When used in a cohort of very low-risk PCa, it was associated with an increased risk of biopsy reclassification due to any unfavourable pathology finding; hence, PSAV RC is proposed for monitoring patients on AS and to decrease the frequency of biopsies in the long term [25].
In summary, higher PSAV and shorter PSADT kinetics might be useful in differentiating between PCa with more aggressive potential. The association between pre-treatment PSAk and PCa biology is supported by multiple studies that found a strong association between the PSAV and PSADT and various pathological features of aggressive PCa [22].

PSA density (PSAD)
The PSA density (PSAD), calculated by dividing the preoperative PSA by the prostate volume (without seminal vesicles), was introduced in the early 1990s by Benson et al. [26]. Its role in predicting upgrading and reclassification in men with low-risk PCa enrolled in AS has been assessed in several studies.
A study recently published by Yusim et al. including 992 men with a median age of 66 years concluded that patients with PSAD higher than 0.34 ng/mL 2 have a 56.4% chance of being diagnosed with a csPCa, with risks estimated at 4%, 8.5% and 31.5%, for PSADs <0.09 ng/mL 2 , between 0.09 and 0.19 ng/mL 2 and between 0. 19  In summary, PSAD is proposed to be tested prior to prostate biopsy as it is an inexpensive and widely available tool, which may avoid unnecessary biopsies. Its inclusion in AS protocols could improve inclusion criteria and follow-up of PCa patients. Furthermore, the more accurate prostate volume calculation by MRI might potentially make its use more reliable [27,29].
Prostate Health Index (PHI) PHI, from Beckman Coulter [30], is a diagnostic blood test that combines free (fPSA), total PSA (tPSA) and isoform [-2]proPSA into a single score, calculated according to the formula [-2]proPSA/fPSA x √ tPSA, developed to maximize specificity at high sensitivity [31]. Approved by the USA Food and Drug Administration (FDA), it is now considered by the NCCN guidelines in the diagnostic setting [4].
Chiu et al., demonstrated that PHI and %p2PSA are predictors of RP pathologic outcomes such as pT3, pGS, GS upgrade, tumour volume >0.5 mL, and Epstein criteria for significant tumours [32]. Moreover, Hirama et al. showed the high diagnostic ability of p2PSA-related parameters, such as %p2PSA and PHI, for discriminating patients with non-significant cancer from those with significant cancer, being useful markers for AS [23].
Several studies compared the predicting value of PHI and PCA3. Cantiello et al., found a superior predictive accuracy of PHI over PCA3 (AUC 0.92 vs. AUC 0.77) in discriminating clinically significant disease in men eligible for AS, outperforming PCA3 performance, resulting in a higher net benefit [33,34].
The use of these markers may reduce the incidence of underestimation at initial diagnosis, enabling a more accurate selection of candidates suitable for AS. PHI measurement could be clinically useful in discriminating the presence of insignificant PCa in AS candidates [23,31-33]. Furthermore, its utility is increased by adding mpMRI to discriminate the presence of pathologically confirmed significant PCa, as demonstrated in a cohort of patients who underwent RP but were eligible for AS [ The 4Kscore ® provides a score ranging from 0 to100%, reflecting the probability of finding a significant PCa at biopsy; hence, 100% minus the 4Kscore result is the personalized negative predictive value (NPV) or probability that a patient will not have Gleason ≥7 cancer on prostate biopsy [37].
Lin et al. evaluated the 4Kscore ® in a cohort of patients candidates for AS, already diagnosed with cancer, demonstrating that the addition of 4Kscore ® to a model containing clinical information significantly improves the prediction of the outcome in the first surveillance biopsy, which is associated with reclassification [38].
Similarly, in a prospective evaluation by Borque-Fernando et al. in an AS scenario, concluded that the 4Kscore ® , at a cut-off 7.5%, was significantly associated with tumour reclassification at the confirmatory biopsy, while the previously used %fPSA/tPSA ratio did not show this association [39].
Due to the ability of the 4Kscore ® to discriminate between men who are likely to harbour clinically relevant PCa and those with indolent tumours or no cancer, it has been proposed as a good marker to identify patients who are more likely to benefit from biopsy, due to a higher risk of csPCa requiring active treatment [39,40] potentially implemented in an AS setting [37-41].
GPS has been shown to predict adverse surgical pathology (AP) and biochemical recurrence (BR) in men diagnosed with low-and intermediate-risk PCa treated with immediate surgery [44,45]. It has been used as a tool to inform the decision making of immediate treatment versus AS in men newly diagnosed with low-or favourable intermediate-risk PCa and was recently included in the NCCN guidelines [46]. Some studies have demonstrated that GPS is associated with an increased risk of AP findings in AS patients who later underwent radical prostatectomy (RP), being also associated with BR following surgery in such patients [46,47]. The GPS ranges from 0 to 100, with higher scores indicating a greater risk of aggressive disease.
Eure et al. reported higher rates of AS uptake (62% vs. 40%) and persistence on AS at 1 year (55% vs. 35%) among men who underwent the GPS testing compared to those who did not [48]. The GPS test has been recommended in men with early-stage PCa [49]. In this way, after obtaining a GPS, more patients were re-classified as very low-risk compared to the initial distribution determined by the NCCN risk category alone [42]. Regarding reproducibility, the initial test is the most informative, and serial testing seems to have limited benefit [43].
Although higher GPS has been associated with an increased risk of AP and BR, it is reasonable to use the GPS in conjunction with other known clinical risk factors when selecting patient candidates for AS [47].
Genome DX Decipher ® Genomic Classifier Decipher ® Prostate Cancer Test, from Genome Dx Biosciences [50], is a 22-feature RNA biomarker assay that has been developed to predict metastasis risk at 5-and 10-years after RP [51]. This test incorporates 22 coding and non-coding genes that cover seven cancer pathways, including angiogenesis, invasion and metastasis, or growth and differentiation [50]. It generates a score ranging from 0 to 1, with higher values indicating an increased probability for both AP and poorer oncologic outcomes [52].
Decipher ® is a significant predictor of AP when used alone or combined with clinical risk stratification systems [53], and when applied to prostatectomy tissue from NCCN very low-/low-and favourable intermediate-risk patients, it has identified patients inappropriately selected for AS.
It has been demonstrated that Decipher ® at the biopsy cores outperformed NCCN clinical risk grouping, biopsy GS and preoperative PSA; for every 10% increase in the Decipher ® score, the high-risk increased by 1.72 [53]. Notably, combining Decipher ® with NCCN risk groups increased the concordance index (C-index) from 0.75 to 0.88, suggesting that Decipher ® captures a considerable proportion of PCa available on diagnostic needle biopsies obtained in routine clinical practice. Decipher ® scores on biopsies were also predictive of other key outcomes, including the presence of primary Gleason pattern 4/5 and an elevated risk of metastasis within 5 years [54]. Applying the Decipher ® test appears to better guide treatment recommendations [55].
The Cell Cycle Progression (CCP) score, from Myriad [56], is a validated prognostic RNA expression signature, based on the expression levels of 31 genes related to cell cycle progression and 15 housekeeping genes. Several studies have demonstrated that the CCP score is strongly associated with progression to metastatic disease after both surgery and radiotherapy [57][58][59].
The CCP score threshold was developed in men who might typically be considered for AS based on having low-or favourable intermediate-risk disease according to a conservative interpretation of NCCN guidelines [46]. The combination of CCP score with clinico-pathologic risk factors from the Cancer of the Prostate Risk Assessment (CAPRA) model [60] results in the cell cycle risk (CCR) test, with a better classification performance. The CCR-based risk stratification might guide the identification of patient candidates to AS from those who would require more active intervention [58]. Both the molecular score and the derived predicted risk might be used in both AA and non-AA patients with newly diagnosed PCa [59].
Lin et al., validated this test in conservatively managed men and showed that the CCR score threshold significantly dichotomized men with high-risk (CCR >0.8) and low-risk (≤0.8) of 10-year PCa mortality, indicating that the threshold can be safely used to identify candidates for AS. Application of the CCR score threshold should result in a substantial increase in men who would be considered candidates for AS that would have traditionally been excluded [57].

Urine Biomarkers
Prostate Cancer Antigen 3 (PCA3) PCA3 was first described in 1999 by Bussemakers et al. [61] as a non-coding mRNA only expressed in human prostate tissue but significantly overexpressed in PCa and is used in clinical practice as Progensa ® PCA3 assay, an in vitro nucleic acid amplification test, developed by Hologic [62], approved by the FDA to be used in conjunction with other patient information to aid the decision for repeat biopsy in men 50 years of age or older with one or more previous negative prostate biopsies.
This PCA3 score is generated as a ratio of PCA3 mRNA to PSA mRNA in urine multiplied by 1000, and a PCA3 score <25 is associated with a decreased likelihood of positive biopsy [63,64].
Tosoian et al., evaluated it in a cohort of men with favourable-risk PCa included in AS programs and who underwent tumour reclassification, demonstrating that patients who reclassified had significantly higher PCA3 scores at both initial (48.0 vs. 24.5, p = 0.007) and subsequent (63.5 vs. 36.0, p = 0.002) measurements, indicating an association between PCA3 and grade reclassification [65].
However, despite the demonstrated PCA3 utility for PCa diagnosis, several studies suggest that additional biomarkers should be incorporated into PCA3 to ensure detection of high grade [66].
The addition of mpMRI to PCA3 reduces overdetection and overtreatment of indolent PCa by improving diagnostic accuracy. Hence, the use of urinary PCA3 testing in men with low or equivocal suspicion mpMRI allows unnecessary biopsies, the refinement of risk stratification and the optimization of high-grade cancer detection to be reduced [64].

TMPRSS2:ERG (T2E) fusion gene
The T2E fusion gene was first discovered in 2005 by Tomlins et al. [67], and it is present in approximately 50% of PCas. It constitutes a highly specific biomarker that can be detected both in FFPE and urine samples [68,69].
In a prospective study by Lin et al., post-digital rectal examination (post-DRE) urine samples were collected and T2E levels were analysed, indicating an association with higher tumour volume and higher GS in subsequent biopsies [70]. Furthermore, high T2E expression levels have been associated with an increased risk of tumour upgrading and upstaging in AS candidates [71].
In an AS context, Whelan et al., assessing the effectiveness of T2E expression in prostatic secretions, showed that non-invasive T2E measurement in urine may refine patient acceptance into AS programs [72].
However, a comparative study between urinary PCA3 and T2E measures during multiple times at surveillance found that these markers add little or no improvement in clinical variables in predicting biopsy reclassification due to the high PCa heterogeneity [73,74]. Hence, both PCA3 and T2E are proposed to be combined in a risk calculator, MiProstate Score (MiPS), which shows higher specificity and sensitivity [75,76].

SelectMDx
Related to urinary biomarkers useful in predicting csPCa, a novel urinary assay-based risk score called SelectMDx from MDXHealth ® was developed by combining serum PSA, PSAD and clinical factors such as age and prior negative biopsy with two mRNA signatures, namely urinary homeobox C6 (HOXC6) and distal-less homeobox 1 (DLX1), and KLK3 gene as a reference [77].
Leyten et al. developed the test based on detecting increased mRNA levels HOXC6, DLX and tudor-domain-containing 1 (TDRD1), which have been selected between eight candidates because of having independent additional predictive value of PSA for the detection of biopsy GS ≥7. These genes have been associated with PCa development [78].
In this study, it was shown that the combination of urinary HOXC6 and DLX1 was superior to Progensa ® PCA3 in the diagnosis of GS ≥ 7 PCa.
Based on the urinary biomarker panel previously defined, Van Neste et al. proposed a new test by combining HOXC6 and DLX1 mRNA expression levels with traditional clinical risk factors, such as PSAD, DRE, PSA, age, history of prostate biopsy and family history, which is able to detect high-grade csPCa accurately and consequently could be used in AS decision making, reducing unnecessary biopsies and potential overtreatment [79].
We have just published our results with SelectMDx and PCA3 in an AS setting, showing that SelectMDx showed statistically significant differences related to pathological progression-free survival (HR: 1.035; 95% CI: 1.012-1.057) (p = 0.002) with a C-index of 0.670 (95% CI: 0.529-0.810) and AUC of 0.714(95% CI: 0.603-0.825) at 5 years. The combination of both biomarkers did not improve the prediction of PP and C-index 0.630 (95% CI: 0.455-0.805) [80]. Despite the high ability of SelectMDx in predicting the likelihood of finding high-grade PCa, it has been demonstrated that its combination with mpMRI could better select candidates to AS, identifying men who harbour csPCa, hence improving the cost-effectiveness. Pepe P et al., found that mpMRI and SelectMDx missed 3/9 (33.3%) and 4/9 (44.5%) of csPCa, respectively. Moreover, mpMRI combined with Select-MDx diagnosed 7/9 (77.8%) csPCa, outperforming SelectMDx alone, in a cohort of men enrolled in an AS protocol [81,82]. Nonetheless, several studies have demonstrated the potential cost-effectiveness of SelectMDx alone and instead of using mpMRI. In Spain, the cost-saving result of using the SelectMDx strategy has been estimated at EUR 247 per patient, which is EUR 20 million per yearly cohort, mostly by preventing the detection of insignificant cancers and, to a lesser extent, by reducing the number of biopsies [83]. Similarly, Dijkstra et al. showed that SelectMDx could improve PCa patients' quality of life and detect high-grade tumours while saving cost compared to the current standard of care [84].

Conclusions
Active Surveillance (AS) is increasingly used in all health systems, balancing screening benefits with lowering overtreatment in PCa. Since this strategy was introduced, AS has evolved from initial protocols a decade ago, allowing inclusion without any reliable imaging, to more refined contemporary management pathways, guided by mpMRI imaging and fusion biopsy. Nevertheless, mpMRI is not widely available everywhere, and it is still very dependent on radiologist expertise and hence lacks homogeneity.
Biomarkers might play a role, probably combined with mpMRI and mainly in PIRADs 3 lesions, the "grey box", where results are unclear. Their potential to give information from the scan the whole gland, their homogeneity, reproducibility and potential for comparison make them very attractive in the AS setting, but their prices should come down in order to be generalized, mainly tissular, markers. Most of them have shown in pilot studies their complementary role with mpMRI, but none is going to replace it when a biopsy is needed, at both the confirmatory and follow-up phases.
The main limitation that we have found in this review is that there are few studies specifically focused on applying these biomarkers in AS series. However, given the applicability that many of these tests have in identifying csPCa, they are good candidates for being tested in an AS scenario (Table 1).
Tissular biomarkers, mainly those analysing independent molecular markers not related to classical clinic-pathological variables, could have a role in the future of AS management. However, issues in terms of high costs and difficulties in handling are drawbacks that should overcome before widespread implementation is possible. Therefore, at present, we can just recommend them in doubtful cases such as high-volume Gleason 3 + 3 or favourable intermediate-risk PCa where a more conservative management could be considered in case of favourable tissular marker scoring. PSA is a glycoprotein secreted by prostatic epithelial cells that lyses the clotted ejaculate to enhance sperm motility. It is the most important commonly used biomarker, but it has been shown not to be a cancer-specific marker, as some prostate diseases could also produce PSA elevated levels Its role as biomarker in AS has been questioned due to its variability and has been mostly studied related to its changes with time, prostate volume or other isoforms [19] PSA kinetics Blood Rate of PSA change over time: PSA doubling time (PSADT) is the number of years over which a certain level of PSA increases by a factor of two and is calculated as DT = ln (2) Can detect high-grade csPCa accurately and could therefore be used in AS decision making, reducing the number of unnecessary prostate biopsies and potential overtreatment [78][79][80][81][82][83][84][85] Prospective and longitudinal studies with different biomarkers focused on AS cohorts are still missing. Prospective, well-designed studies assessing predefined clinically relevant endpoints are needed to fully assess the real potential of these biomarkers and to compare them with other diagnostic tools, such as novel predictors of grade progression and mpMRI-PRECISE criteria [85]. Robust clinical evidence derived from such studies will support and guide clinical decision-making in the selection, management and long-term follow-up of PCa patients in AS schemes.   [CrossRef] 42. About the Oncotype DX Genomic Prostate Score ® Test Oncotype IQ ® . Available online: https://www.oncotypeiq.com/en-US/ prostate-cancer/healthcare-professionals/oncotype-dx-genomic-prostate-score/about-the-test (accessed on 5 February 2021).