FormalPara Key Points

Proclarix is a new blood-based marker test based on the combination of different glycoproteins providing a risk score of clinically significant prostate cancer.

In men with prostate-specific antigen (PSA) levels between 2 and 10 ng/mL, prostate volume ≥ 35 cm3, and normal digital rectal examination, Proclarix can reach up to 90% sensitivity for clinically significant prostate cancer.

Initial studies have demonstrated that Proclarix could be integrated in the diagnostic algorithm of prostate cancer in order to reduce unnecessary prostate biopsies.

1 Introduction

Prostate cancer (PCa) is the second most commonly diagnosed malignant neoplasm in men, representing 15% of all cancers diagnosed worldwide [1]. Suspicion of PCa continues to be based on serum determination of prostate-specific antigen (PSA) and digital rectal examination (DRE). However, diagnosis of the neoplasm requires its demonstration by a prostate biopsy. This procedure has a low specificity, which has led to an excessive number of unnecessary biopsies and an overdetection of insignificant PCa [2]. The PSA is a specific protein of the prostatic tissue, therefore its serum concentration also increases in benign processes that affect the prostate gland, such as hyperplasia or inflammation [3, 4].

In order to increase the specificity of PSA, multiple parameters have been used, such as PSA density (PSAD) [5], age-related ranges of PSA [6, 7], percentage of free PSA (%fPSA) [8], and more recently, through some markers determined in blood or urine. Among all urinary biomarkers, the most relevant are PCA3 [9], SelectMDx, and ExoDx [10], while in serum there are the Prostate Health Index (PHI) [11] and 4K test [6, 12, 13]. These are used to improve specificity and therefore reduce the number of unnecessary biopsies [14].

Multiparametric magnetic resonance imaging (mpMRI) allows the detection of suspicious PCa lesions and establish the risk of clinically significant PCa (csPCa) through the Prostate Imaging-Reporting and Data System (PI-RADS) [15]. MpMRI can substantially increase the specificity of PSA to reach a negative predictive value (NPV) near 90% [16]. Moreover, it allows MRI-targeted biopsies to suspicious areas to be performed, increasing the sensitivity for csPCa [17, 18].

Nowadays, we are faced with substantial improvement in the csPCa diagnostic strategy, with new criteria that are mainly based on the determination of PSA [19]. Furthermore, there is significant progress in the efficacy of prostate biopsies, based on the association of targeted and systematic biopsies [20]. However, new tests are being developed that can still improve the csPCa diagnostic strategy, such as predictive models based on mpMRI and other clinically independent variables, or the use of new biomarkers.

In this context, the new test named Proclarix (Proteomedix, Switzerland) has recently been introduced. It is a csPCa risk score based on the serum determination of thrombospondin-1 (THBS1), cathepsin D (CTSD), PSA, and %fPSA, together with age. Proclarix has recently been approved by the Conformité Européenne (CE) in order to be used in men with PSA levels between 2 and 10 ng/dL, normal DRE, and prostate volume (PV) ≥ 35 cm3. The main objective of this systematic review is to analyze the current clinical usefulness of Proclarix for csPCa, posterior to the study of the bases and development of Proclarix.

1.1 Bases and Development of Proclarix

In 2010, Cima et al. [21] identified multiple promising biomarkers to potentially improve the diagnosis of csPCa. The glycosylated proteome of a mouse model of progressive PCa caused by the inactivation of prostate-specific phosphatase and tensin homolog (PTEN) was the basis of their study. Simultaneously, Surinova et al. [22] highlighted the importance of proteomics and technological requirements in the search for new biomarkers.

Kälin et al. [23] focused on the inactivation of the PTEN pathway to discover new biomarkers related to the prognosis of patients with metastatic castration-resistant PCa. THBS1 was identified as a marker for distinguishing benign from malignant prostate disease.

In 2017, Endt et al. [24] evaluated the role of four glycoproteins—CTSD, intercellular adhesion molecule-1 (ICAM-1), olfactomedin-4 (OLFM4), and THBS1—in the early diagnosis of PCa. The function of these glycoproteins is based on the characteristics of cancer development, i.e. apoptosis, angiogenesis, and metastasis. OLFM4 was shown to be an important regulator of apoptosis in murine PCa cells [25], while THBS1 has been reported to be a regulator of angiogenesis in malignant and non-malignant prostate tissue [26]. Furthermore, it was observed that both CTSD and ICAM-1 promoted the malignancy of prostate epithelium [27]. In that study, only THBS1 and %fPSA were significantly different in univariate analysis between the PCa-positive and -negative groups. CSTD was shown to have a statistically significant additional benefit when combined with THBS1, while ICAM-1 and OFLM2 did not. In this way, Kälin et al. [23] provided evidence that the measurement of CTSD and THBS1 in combination with %fPSA and age, improved the diagnosis of csPCa. In the same year, Tennstedt et al. [28], through a retrospective study that included patients with elevated PSA, normal DRE, and elevated PV, analyzed the role of CTSD, ICAM-1, THBS1, OLFM4, tissue inhibitor of matrix metalloproteinase-1 (TIMP-1), and hypoxia upregulated protein-1 (HYOU1), and concluded that in a population with these characteristics, the combination of CTSD and THBS1 was significantly more accurate than %fPSA in determining the absence of PCa.

In 2020, Klocker et al. [6] evaluated a new diagnostic test named Proclarix, which incorporated THBS1 and CTSD along with total PSA (tPSA), %fPSA, and age, to predict csPCa. In the analyzed study, the test showed higher diagnostic accuracy compared with %fPSA alone. Therefore, Klocker et al. concluded that Proclarix achieved a high clinical performance, related to a greater specificity which could help decide which subjects with suspected csPCa should undergo a prostate biopsy.

On the other hand, in 2020, Macagno et al. [29] evaluated the stability of THBS1 and CTSD at different temperatures before and after storage. They demonstrated that the results were unaffected by storing the serum samples at 2–8 ºC for up to 2 weeks and by freezing them at – 20 ºC and subsequent thawing before measurements. They also studied the coefficient of variation (CV) of replicate measurements corresponding to the repeatability of the test, obtaining a rating of 5.5% for THBS1 and 4.3% for CTSD. Reproducibility was analyzed by measuring the same samples in two different laboratories, obtaining an estimated overall bias of − 6.3 for THBS1 and − 3% for CTSD. However, these results differ by < 20% independent of the lot used and the laboratory performing the assays. This way, Macagno et al. were able to confirm that the ELISA technique could be used in the in vitro diagnosis of Proclarix and that the proteins were suitable for use in routine clinical practice.

2 Evidence Acquisition

2.1 Search Strategy and Results

A bibliographic search in the PubMed, Cochrane, and Trip databases was conducted by two independent authors (MC and JM) in January 2022 using the Medical Subject Heading (MeSH) [30] terms ‘thrombospondin-1’, ‘cathepsin-D’, ‘Proclarix’, and ‘prostate’. All studies written in English were considered. In order to include studies that analyzed the same combination of biomarkers as Proclarix, despite not mentioning its commercial name, the following Boolean operators were used: (((Thrombospondin 1) OR (Cathepsin D)) OR (Proclarix)) AND (Prostate). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [31] statement was followed and the Population, Intervention, Comparison and Outcomes (PICO) selection criteria were established as men with suspected PCa undergoing prostate biopsy, comparing THBS1 and CTSD with %fPSA and analyzing the biopsy results [32].

A total of 181 reports that included any of the two biomarkers or Proclarix and mentioned the keyword ‘prostate’ were detected in PubMed; one duplicate article was excluded. By reading the title and abstract of these studies, 176 were discarded. Twenty-eight articles were excluded for evaluating CTSD or TBHS1 in other pathologies or in animals, 67 investigated the physiological processes involving both molecules, 29 evaluated the correlation between both biomarkers and the risk of progression, 42 studied the THBS1 or CTSD role in angiogenesis, 9 evaluated the relationship between THBS1 and benign prostatic hyperplasia and one analyzed the stability of both molecules for clinical practice. Finally, four articles that met the PICO selection criteria were included. The PRISMA flowchart of study selection is summarized in Fig. 1.

Fig. 1
figure 1

PRISMA flowchart of study selection

Full size image

In order to find ongoing and unpublished articles, the International Clinical Trials Registry Platform (ICTRP) and ClinicalTrials.gov were employed. Three articles were obtained, one of which was a duplicated observational study [33] and the remaining two were clinical trials that have not yet been published.

Based on their title and abstract, articles were selected for full-text review according to the relevance of the issue of interest. Only studies analyzing biomarkers in PCa were included, while those that focused on THBS1 and CTSD but in different pathologies were excluded. Finally, following the PICO criteria, four original articles were selected.

2.2 Quality Assessment

The Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) [33] tool was used to analyze the quality of the included studies, and the results are presented in Table 1 and Fig. 2. No information was provided regarding the blinding of the collected samples in the Steuber et al. [35] publication nor in the Morote et al. article [36], which is why the risk of bias due to the index test in both studies was unclear. The remaining studies clearly specify how the patient selection was made, the values of the index test, the reference standard, and the flow and timing of the data.

Table 1 Quality assessment results of the selected studies according to QUADAS-2
Full size table
Fig. 2
figure 2

Risk-of-bias assessment according to QUADAS-2. QUADAS-2 Quality Assessment of Diagnostic Accuracy Studies 2

Full size image

2.3 Data Extraction

The selected studies were thoroughly examined and the following information was extracted from each article: name of first author, year of publication, type of study, sample size, inclusion and exclusion criteria, definition of csPCa, diagnostic approach, endpoint, and results (Table 2).

Table 2 Descriptive characteristics and results of selected studies
Full size table

3 Evidence Synthesis

Four studies were identified that met the PICO criteria established in this systematic review. Steuber et al. [34] conducted a single-center retrospective study including men with a PSA level between 2 and 10 ng/mL, normal DRE, and PV ​≥ 35cm3. Serum samples from 474 men were obtained prior to the biopsy, stored at room temperature for 30 min, and centrifuged at 2800g for 10 min in a serum separator tube. Subsequently, samples were kept at – 80 ºC for long-term storage. Patients underwent 10-core TRUS-guided prostate biopsy but no mpMRI data were available. Of the 474 serum samples used, 67 were analyzed in a previous study [35]. Moreover, all men with a positive biopsy underwent a subsequent radical prostatectomy (RP). The cohort was evenly divided into training and validation sets according to the cut-off date. With a sensitivity of 90%, Steuber et al. compared the combination of THBS1, CTSD and %fPSA with %fPSA alone, obtaining a specificity of 62% and 22.5%, respectively. Thus, they determined that the use of %fPSA alone avoided 24% of biopsies, while for the combination of the three variables, the rate was 62%. On the other hand, csPCa was defined as a Gleason score (GS) ≥ 7. The combination of THBS1, CTSD and %fPSA failed to detect a total of 12 patients with a GS ≥ 7 and three cases of GS ≥4 + 3 (prostatectomy specimen grading). A better correlation between the GS from prostatectomy and that obtained from biopsy was found. It is important to take this into consideration, since upgrading a GS from biopsy to prostatectomy occurs frequently.

Klocker et al. [6] evaluated the named Proclarix test with a retrospective study using samples from two centers. One of the centers (comprising 474 patients) was also analyzed in the study by Steuber et al. [34]. Klocker et al. studied a total of 955 men with PSA levels between 2 and 10 ng/mL, normal DRE, and PV ≥ 35cm3 determined by TRUS. Patients underwent TRUS-guided 10- to 12-core prostate biopsies. CsPCa was defined as an International Society of Urological Pathology (ISUP) grade group (GG) ≥ 2. Using a risk score cut-off at 10%, the sensitivity of Proclarix was 90%, with 43% specificity and 95% NPV. This is in comparison with %fPSA alone, which, with the same sensitivity, resulted in a specificity of 17% and an NPV of 89%. No GG 5 were missed and 10% of csPCa were not detected. The risk score calculated by Proclarix increased across groups and could thus differentiate the aggressiveness of csPCa detected on biopsy. The number of avoided biopsies was more than double in the biomarker model (37%) compared with %fPSA (16%), while keeping the number of missed cancers constant.

The study by Steuber et al. [33], published in 2021, analyzed the diagnostic accuracy of Proclarix in patients with suspected PCa. CsPCa was defined as an ISUP GG ≥2 detected on biopsy. The patients included in the study had PSA levels between 2 and 10 ng/mL, normal DRE, and PV ≥ 35 cm3. Men underwent TRUS-guided 10- to 12-core biopsies. Furthermore, since its inclusion in the recommendations in the European Association of Urology (EAU) guidelines in March 2019, additional mpMRI-fusion biopsies were performed in those patients with a positive result on mpMRI (PI-RADS scores of 3–5). The study population consisted of 362 men who met the inclusion criteria and underwent a TRUS-guided biopsy, of whom 121 obtained a positive mpMRI result and underwent a fusion biopsy. The serum samples were collected before the prostate biopsy, were centrifuged at 2500g and frozen at – 20 ºC. Laboratory technicians were blinded to the clinical and pathological information. At equal sensitivity, Proclarix had significantly higher specificity (22% vs. 14%) compared with %fPSA alone. In the mpMRI-fusion biopsy cohort, the Proclarix density was calculated, taking the PV into account. Thus, Steuber et al. found that with a sensitivity of 97%, Proclarix density was more specific than PSAD (33% vs. 8%; p < 0.001), being able to avoid up to one-third of unnecessary prostate biopsies. The risk score was significantly correlated with aggressiveness of PCa detected on biopsy. In the total cohort, six cases would have been missed by %fPSA and Proclarix, but with different aggressiveness: four GG 2, one GG 4, and one GG 5 with %fPSA, compared with five GG 2 and one GG 3 with Proclarix. Both Proclarix density and PSAD missed one GG 2 in the mpMRI subcohort. In men with an mpMRI, Proclarix density accurately diagnosed 31 of 32 patients with csPCa and correctly ruled out 27 of 66 men without cancer.

Morote et al. [36] analyzed the efficacy of Proclarix in the selection of candidates for mpMRI and derived prostate biopsies among men with suspected PCa. A retrospective study in a prospective database was carried out analyzing 567 patients with suspected PCa after mpMRI. Overall, 286 men had serum PSA levels between 2 and 10 ng/mL, normal DRE, and PV ≥ 35 cm3 (Subset 1), while 286 men had a PSA outside the range, an abnormal DRE, or a PV <35cm3 (Subset 2). Samples were obtained before prostate biopsy and stored at – 80 ºC. CsPCa was defined as an ISUP GG ≥ 2 detected on biopsy. Twelve-core systematic biopsies were performed in all patients and in those with PI-RADSv2.0 lesions ≥ 3; an additional two to three guided biopsies were obtained from each lesion. A correlation between PCa grading and Proclarix was found, with similar results obtained as Steuber et al. [36]. However, Morote et al. also found an association between Proclarix and PCa clinical stage and the risk of recurrence of treated localized PCa. Analyzing the test in the population with specific characteristics, they obtained a sensitivity of 95.8% and a specificity of 32.5%. Nonetheless, in the remaining patients, the sensitivity of Proclarix was similar (98.1%) but with lower specificity (17.2%), thus being able to avoid a higher number of mpMRI and derived prostate biopsies in Subset 1 (25.3% and 8.7%, respectively) with 4.2% and 1.9% misdiagnosis rates.

4 Discussion

Although the combination of DRE and PSA allows for correct PCa risk stratification, a significant number of patients in this challenging clinical setting leave the physician uncertain as to whether to perform a prostate biopsy or not due to the low specificity. For this reason, in the last several years, numerous biomarker-based tests have been developed in order to help in the early detection of PCa. Different tests are available to predict the presence of csPCa and to help clinical decision making on who to biopsy and who to re-biopsy after an initially negative biopsy result.

In this systematic review, we analyzed four articles that met the PICO selection criteria to evaluate Proclarix in the clinical setting. The test includes serum values of THBS1 and CTSD, as well as patient age, tPSA and %fPSA, in order to discriminate csPCa from other prostate conditions [6].

All articles had a low risk of bias, evaluated using the QUADAS-2 tool, even though three of the studies were retrospective in nature. The population in all studies was similar, with the same inclusion criteria, i.e. patients with PSA levels between 2 and 10 ng/mL, normal DRE, and PV ​≥ 35 cm3, with the exception of the study by Morote et al. [36], who also included a subgroup of patients who did not meet any of these conditions. The collection of samples was homogeneous in all studies. All studies performed 10- to 12-core TRUS-guided prostate biopsies to confirm the diagnosis, and in those patients in the Steuber et al. [33] and Morote et al. [36] studies with an mpMRI, 2- to 3-core guided biopsies were obtained from each PI-RADSv2.0 lesion ​≥ 2. All studies concluded that the combination of THBS1, CTSD, patient age, and %fPSA had a higher specificity when compared with %fPSA alone. Proclarix demonstrated the ability to avoid up to 62% of unneeded biopsies [34], keeping the number of missed csPCa constant. In this context, this could be useful when deciding which patients could safely forgo a biopsy. Table 3 summarizes the performance characteristics of Proclarix in each study.

Table 3 Performance of Proclarix in published studies
Full size table

The definition of csPCa is a dynamic process that has been changing throughout the years. In 1994, Epstein et al. published the first criteria to define csPCa, and developed a prediction model to determine which patients would not need definitive therapy based on the prostatectomy specimen [37]. They considered a clinically insignificant PCa when the tumor volume was < 0.2 cm3, no Gleason pattern of 4 or 5, and there was no seminal vesicle or lymph node invasion. In 2011, Ahmed et al. [38] proposed two definitions for csPCa based on three biopsy parameters. First, they included the total cancer core length (TCCL) with values ≥ 10 mm and ≥ 6 mm depending on individual preference, comorbidity, age, and life expectancy. They also incorporated two lesion volume thresholds measured using the maximum cancer core length (MCCL) ≥6 mm and ≥4 mm. Finally, they combined dominant and non-dominant Gleason pattern 4 using the csPCa definition. Thus, Ahmed et al. concluded that analyzing the prostate biopsy sample could be useful for increasing the proportion of men who choose, or are advised, to undergo active surveillance while also ensuring that those who require therapy do actually undergo the therapy. Three articles [6, 33, 36] used the ISUP definition in order to classify PCa, and considered csPCa a GG ≥ 2, while Steuber et al. [34] defined csPCa as a GS ≥ 7. GS upgrading, in which GS obtained from RP is higher than that obtained from biopsy, is consistently reported as prevalent in cohorts from around the world. In their review, Alchin et al. [39] described that the GS upgrade after radical prostatectomy could go from 29.6% to 45.6%, depending on the study. This can be predictive of subsequent biochemical recurrence and oncological failure. Steuber et al. described a better correlation between GS obtained from prostatectomy than that obtained from biopsy [34]. Obtaining less discordances could potentially have an impact on the choice of treatment and even on oncological outcomes.

In the majority of studies carried out to date, Proclarix has been evaluated in patients with specific characteristics (PSA levels between 2 and 10 ng/mL, normal DRE, and PV ≥ 35 cm3). Only Morote et al. [36] analyzed a subset of 286 patients who did not meet these criteria and obtained a sensitivity of 98.1%, a reduction of unnecessary biopsies of 8.7%, and a misdiagnosis of csPCa of 1.9%; however, this was a retrospective design in one center and more studies are needed in order to validate these results.

The introduction of mpMRI and fusion-guided biopsies have improved the early detection of csPCa [41]. Steuber et al. [34] included the use of mpMRI for fusion biopsy in 121 patients and concluded that the performance of Proclarix improved when used in conjunction with mpMRI in the decision to biopsy the patient. Moreover, in this subgroup of patients, when PV was taken into account, measuring Proclarix density, the NPV and specificity further increased 97% and 33%, respectively, which is why Steuber et al. suggested taking PV into account to calculate Proclarix density for men undergoing mpMRI. On the other hand, Morote et al. [36] also analyzed the role of Proclarix in patients with suspected PCa after mpMRI and concluded that Proclarix was correlated with clinical stage, with lower values in those patients with localized tumors and higher in disseminated PCa. Furthermore, Proclarix was also associated with the risk of recurrence of treated localized PCa.

Three studies used a retrospective study design, while the PROPOSe study [33] used an observational prospective design. The risk of bias was evaluated using the QUADAS-2 tool. All articles provided a clear definition of the inclusion and exclusion criteria, and collected the samples consecutively, having a low risk of bias in the patient selection domain. No information was provided regarding the blinding of the collected samples in the studies by Steuber et al. [35], therefore the bias in the index test domain was unclear. The remaining studies clearly specify the index test values, the reference standard, and the flow and timing of the data.

There are some established risk factors for total PCa incidence, such as older age, African American race, and positive family history of PCa [41], however none of the studies included these factors in the Proclarix prediction model. In future studies, the role of Proclarix in these patients should be analyzed in order to determine whether the sensitivity, specificity, and number of avoided prostate biopsies are the same.

It is difficult to compare Proclarix with other commercially available tests since there are no head-to-head studies analyzing these tests. In 266 biopsy-naïve men who underwent mpMRI, performing the 4K score test would avoid 12% of unnecessary biopsies misdiagnosing 1.4% of csPCa [42]. Wagaskar et al. developed a 4K score/mpMRI-based nomogram to predict csPCa and to avoid unnecessary biopsies [42, 43]. In a prospective multicenter study including 310 biopsy-naïve men, SelectMDx avoided 37% of unnecessary biopsies, with 10% of undetected csPCa [44]; however, the test seems to be more sensitive than mpMRI but less specific [33]. The PHI was compared with PSAD in an observational study including 232 men scheduled for prostate biopsy [44]. Steuber et al. concluded that PSAD performed better than PHI in csPCa detection and was able to spare more biopsies [33]. Proclarix showed the highest sensitivity for detecting csPCa, which could be used to select appropriate candidates for mpMRI and prostate biopsies [33]. Nonetheless, the results obtained must be analyzed in a cost-benefit study in order to compare all biomarkers and determine the appropriate time to use them in the diagnostic algorithm of PCa. Table 4 shows the comparison results of different commercially available tests.

Table 4 Comparison of four available tests for csPCa detection in different studies
Full size table

The use of Proclarix before and after mpMRI has not yet been studied; however, two clinical trials are still to be completed (MULTI-IMPROD, ClinicalTrials,gov identifier NCT02241122; and ProBioM, ClinicalTrials.gov identifier NCT03730324) that include the use of mpMRI in the diagnostic algorithm for PCa. A cost-benefit study of Proclarix remains to be carried out to determine the most appropriate time to use the test in order to improve the diagnostic algorithm of PCa.

5 Conclusion

Proclarix is a new blood CE-marker test for csPCa detection, based on the serum measurement of THBS1, CTSD, total PSA, and %fPSA, as well as age. Proclarix has been developed in men with suspected PCa and with serum PSA levels between 2 and 10 ng/mL, normal DRE, and PV​ ≥ 35 cm3. This review corroborates that Proclarix achieved a higher level of clinical performance than %fPSA in detecting csPCa in the classical pathway for diagnosis of PCa. However, the current work-up of csPCa has changed after the recommendation of the European guidelines to perform mpMRI and fusion-guided biopsies. It is necessary to demonstrate that Proclarix can be helpful among men with suspected PCa with serum PSA outside the 2–10 ng/mL range, abnormal DRE, and PV < 35 cm3, since only one study has been conducted to date. Moreover, Proclarix must be analyzed in the setting of men with suspected PCa, before and after mpMRI, to select appropriate candidates for mpMRI or prostate biopsy. Finally, cost-effective studies should confirm the efficiency of Proclarix.