sE-cadherin serves as a diagnostic and predictive parameter in prostate cancer patients

Estimating PCa malignancy in an era of personalized healthcare requires novel diagnostic and prognostic biomarkers to develop individual treatment strategies. Therefore, the potential of a panel of serum proteins to detect PCa and predict histopathological outcome in curatively treated disease was evaluated.

Out of 119 candidate molecules, sE-cadherin yielded the most promising potential to serve as a biomarker in the current study cohort with almost 96 % clinically significant PCas according to the Epstein criteria [16]. E-cadherin, a type-I-cadherin, belongs to the cadherin family of transmembrane or membrane-associated glycoproteins located in the adherens junction and basolateral membrane in epithelial cells. It consists of a large extracellular domain, a transmembrane segment and a conserved cytoplasmic domain [18]. E-cadherin is indispensable in promoting cell-cell adhesion in a Ca²⁺-dependant manner and steers tissue morphogenesis [19]. During cancer progression, transcriptional E-cadherin reprogramming induces decreased adhesion and enhanced migration and invasion during the epithelial-to-mesenchymal transition of epithelial cells [20, 21]. sE-cadherin is an 80-kDa ectodomain of the full-length E-cadherin, which is proteolytically shed into the extracellular space through cleavage by secretases and caspases [18]. Concurrent to the disruption of adhesion junctions, EGFR and Wnt/β-catenin pathway signaling, often involved in tumorigenesis, are activated [22]. Since E-cadherin cleavage has been linked to neoplastic adenoma-cancer progression, serum levels of sE-cadherin have been shown to be augmented in patients with breast, gastric and colorectal cancer [23]. This is in line with our findings demonstrating that the serum concentration of sE-cadherin was elevated in patients with PCa, more than other investigated candidates. Thanks to its stability in blood, sE-cadherin may be a suitable indicator for early PCa detection, due to accumulation during tumor-associated proteolysis [23].

So far, little has been reported about sE-cadherin in connection with PCa. Ahmed et al. [24] investigated its potential as a biomarker in a group of 71 Egyptian males with PCa. In agreement with our findings, significant over-expression of sE-cadherin in the serum of tumor patients, compared to healthy individuals, was observed. No correlation with the clinical tumor stage was apparent. In contrast to our findings, the authors reported high expression of sE-cadherin, correlating with a poor Gleason grade and serum PSA. It is difficult to interpret their data in the context of our findings due to the limited sample size of that cohort, median PSA of 70.7 ng/ml for tumor patients, most cancers classified as stage D and details about their staging procedure absent. These study characteristics harbor significant probability of undetected metastasis and therefore deviate from the present study population, which was clinically free from visceral- and bone-disease. Furthermore, alterations in Gleason grading procedures over time have resulted in significant tumor up-grading [25], making a study published in 1999 [24] difficult to compare with the present study carried out more than a decade later, due to bias stemming from the Will Rogers phenomenon. Later, Kuefer et al. [26] observed no difference between the expression of serum sE-cadherin in 61 men with localized PCa compared to controls. On the contrary, a more contemporary series also involving 61 patients with clinically localized PCa, reported by Iacopino et al. [27], demonstrated serum sE-cadherin over-expression in PCa patients compared to that of healthy individuals, which is similar to our observation. However, no correlation of the marker with PSA, T-stage, GS, grade or R- and Pn-status was observed by those investigators. To best knowledge, the current assessment of sE-cadherin has taken place in the largest cohort of clinically non-metastatic males with PCa. sE-cadherin correlated with the highest Gleason pattern, which is an important prognosis factor [28] and was also associated with a Gleason upgrade and pGS, which is more accurate than bGS at predicting biochemical recurrence after RPE [29]. Most importantly, for identifying more aggressive tumors with GS ≥ 7 and tumors with a Gleason upgrade, sE-cadherin outmatched serum PSA without correlating with this marker.

The negative serum sE-cadherin correlation with more aggressive PCa in this investigation requires further research since serum sE-cadherin has been reported to rise with more aggressive or advanced stages in many other cancer types [23]. However, Juhasz et al. has provided evidence for significantly decreased expression of serum sE-cadherin in individuals with advanced and metastasized diffuse-type gastric cancer in contrast to intestinal cancer [30]. Careful analysis of cancer type as well as stage and grade specific molecular machinery pertaining to E-cadherin turnover is therefore warranted. Ongoing studies on sE-cadherin with respect to PCa should also take into consideration that proinflammatory cytokines [31], growth factors [32], as well as chemically induced oxidative stress [33] may contribute to its generation and possibly falsify results based on its expression.

Although sE-cadherin seems to be a promising biomarker, it may not be alone. Catenins, as well as CD44, are associated with E-cadherin in regulating prostate cancer cell adhesion [34]. In particular, CD44v8–10 have been shown to be negatively correlated with c-Myc expression, both in vitro and in vivo. The negative correlation may partly be attributable to redox stress-enhanced Wnt/beta-catenin signaling [35].

The reason no distinct difference was apparent between the CD44 serum level in healthy subjects and tumor patients may have been because CD44 variants such as CD44v5, CD44v6 and CD44v10, possibly involved in prostate cancer progression, were not individually evaluated [36,37]. EpCAM differences between tumor patients and controls were also not apparent in the present investigation, which is in line with other reports [38]. EpCAM has been reported to be responsible for persistence of minimal residual disease and latent relapse of prostate cancer, presumably by affecting the susceptibility to the EGF ligand and regulating the AMPK signaling pathway [39]. Nevertheless, a prognostic/diagnostic role of EpCAM in EGFR activated prostate cancer cells should not be ruled out.

The same ambivalence might apply to MMP9. Although this protein has been suggested to correlate with prostate cancer progression, the correlation has not always been substantiated. Recent experiments documented that MMP9 might even exert tumor-suppressive properties. In fact, elevated MMP9 secretion has been shown to result in decreased prostate cancer cell proliferation, coupled to increased sE-cadherin shedding [40]. The combined effect might explain the inverse correlation between sE-cadherin and MMP9 found in the present investigation, whereby reduced MMP9 and elevated sE-cadherin may drive tumor proliferation forward. Still, this is hypothetical and requires further evaluation.

The present investigation is limited by its single-center design and lack of external validation. To account for regional and racial disparity regarding the biological characteristics of PCa [41], validation studies with patient cohorts from other centers would be desirable. The potential of sE-cadherin as a biomarker could thereby be more exactly appraised.