Clinical implications in the shift of syndecan-1 expression from the cell membrane to the cytoplasm in bladder cancer

Voided urine samples were centrifuged to separate the supernatant from the cellular pellet. The supernatant was decanted and aliquoted, and the urinary pellet was snap frozen. Both the supernatant and pellet were stored at -80°C prior to analysis. Urine supernatant protein concentration was determined using Pierce 660-nm Protein Assay Kit (Thermo Fisher Scientific Inc., Waltham, MA, USA). The level of human SDC-1 (Cat# ab46507 Abcam, Cambridge, MA, USA) was monitored in urine samples using a commercial ELISA assay. The assay was conducted according to the manufacturer’s instructions. A calibration curve was prepared using purified standards for SDC-1. Curve fitting was accomplished by either linear or four-parameter logistic regression following manufacturer’s instructions. Laboratory personnel were blinded to final diagnosis.

Under Institutional Review Board approval with a waiver of consent, 185 bladder tumor paraffin blocks and eight benign bladder paraffin blocks dating from 2002-2009 were identified in the pathologic archives of Orlando Health Department of Pathology. The eight benign bladder paraffin blocks were from autopsy cases in which there was no record of BCa, hematuria or tobacco use. Median follow-up of the patients was 18 months. All paraffin blocks were examined by H&E for histological verification of urothelial carcinoma only histology. Paraffin blocks were cut 5 μm sections and placed on a Superfrost Plus Microslide. Sections were deparaffinized followed by antigen retrieval using citric acid buffer (pH 6.0, 95°C for 20 min). Slides were treated with 1% hydrogen peroxide in methanol to block endogenous peroxidase activity. After 20 min blocking in phosphate buffered saline (PBS) containing 1% bovine serum albumin (BSA), slides were incubated overnight at 4°C with anti-human SDC-1 antibody (mouse monoclonal–Abcam ab34164, dilution 1/400 in PBS containing 1% BSA). Next, slides were incubated with 2 μg/mL of biotinylated anti-mouse IgG secondary antibody (Vector Laboratories, Burlingame, CA) for 30 min at room temperature. Subsequently, the sections were stained using Standard Ultra-Sensitive ABC Peroxidase Staining kit (Pierce/Thermo Fisher Scientific, San Jose, CA) and 3, 3′-diaminobenzidine (DAB; Vector Laboratories), counterstained by hematoxylin, dehydrated, and mounted with a cover slide. Human liver tissue, known to stain strongly for SDC-1, was used as a positive control and omitting the primary antibody served as the negative control. The above immunostaining for SDC-1 as well as the interpretation of the immunostaining for SDC-1 were based on a previous report by Mukunyadzi, et al. [12]. Briefly, the location of immunoreactivity (e.g., nuclear, cytoplasm, cell membrane, and stroma) was noted. The sections were analyzed and staining assessed using a semiquantitative grading system as follows: negative (-), complete lack of staining or staining in <10% of tumor cells; weak (+), staining in 10 to 20% of tumor cells; mild (++), staining in 20 to 50% of tumor cells; moderate (+++), staining in 50 to 70% of tumor cells and; strong (++++), staining in >70% of tumor cells. Using light microscopy, two investigators (MM and AL), who were both blinded to patients’ data, interpreted immunostaining results. A third investigator (CJR) reviewed discrepancies and rendered a final score.

The Wilcoxon rank sum test was used to determine the association between urinary SDC-1 and BCa. Nonparametric receiver operating characteristic (ROC) curves were plotted and the ability of the urinary SDC-1 biomarker to indicate BCa was estimated by calculating the area under the ROC curves (AUC). The sensitivity and specificity of the biomarker at the optimal cutoff value was defined by calculating the Youden index [13]. The agreement between interpreting SDC-1 immunohistochemistry results between the two investigators was analyzed using kappa statistics with the strength of agreement 0.81-1.00 interpreted as almost perfect. The results are presented as weighted kappa with 95% confidence interval (CI). Comparison of immunohistochemical distribution data was performed using Chi square test. Disease-specific survival (DSS) curves were obtained using the Kaplan-Meier method, and compared by the log rank test for each prognostic variable [14]. Multivariate analysis was performed to identify independent prognostic variables using a stepwise Cox proportional hazards regression model. Statistical significance in this study was set at p < 0.05 and all reported p values were 2-sided. All analyses were performed using SAS software version 5.00 (San Diego, CA).

Characteristics of the study cohort of 308 subjects (102 subjects with active BCa and 206 subjects with no evidence of active BCa or a history of BCa) are presented in the Table 1. The median urinary concentration of SDC-1 was not significantly higher overall in subjects with BCa compared to subjects without BCa (71.25 ng/ml vs. 36.10 ng/ml, p = 0.23) (Figure 1). Neither did SDC-1 levels differ among the groups that made up the diverse control cohort (p = 0.562, data not shown). However, a difference in urinary SDC-1 level was noted between patients with tumors of differing grade and invasive subtype. Specifically, low-grade bladder tumors were noted to have higher median urinary SDC-1 levels compared to high-grade bladder tumors (64.55 ng/ml vs. 26.1 ng/ml, p < 0.0001), and non-muscle invasive bladder cancer (NMIBC) had higher median urinary SDC-1 levels compared to muscle invasive bladder cancer (MIBC) (58.23 ng/ml vs. 28.53 ng/ml, p = 0.0049) (Figure 1).

Characteristics of the study cohort of 193 subjects (185 subjects with urothelia carcinoma histology only and 8 subjects with benign bladder histology) are presented in the Table 1. The pathologists’ intra-observer agreement on SDC-1 interpretation and scoring was ‘good’ with a noted kappa score of 0.64 (0.8–1.0, excellent; 0.6–0.8, good; 0.4–0.6, moderate; 0–0.4, poor), 95% CI 0.61–0.68. The percentage agreement was 82.0%. In normal tissue, as well as low-grade disease and NMIBC, over 70% of SDC-1 immunostaining was located within the cellular membrane (Figure 2a) and was graded as moderate (+++) to strong (++++). Minimal immunoreactivity was noted in the stroma. Within bladder tumors, 55% of high-grade tumors (compared to low-grade tumors, p < 0.0001) were noted to have increased cytoplasmic expression and reduced membranous expression of SDC-1 (Figure 2b). In the same way, higher stage tumors (T2-4 vs. Ta-1, p < 0.0001) were noted to have increased cytoplasmic expression and reduced membranous expression of SDC-1 (Figure 2c). Though the location of staining changed from membranous to cytoplasmic amongst high-grade and high stage tumors, immunostaining grading, weak (+) to strong (++++), did not change, illustrating a shift of the ubiquitously expressed SDC-1 from the cellular membrane in well-differentiated, low stage tumors to the cytoplasm in poorly-differentiated, higher stage tumors.

Univariate analysis revealed that NMIBC and membranous immnostaining for SDC-1 represent favourable prognostic factors associated with disease-specific survival (DSS) (p < 0.0001 and p = 0.0004, respectively) (Figure 3). However on multivariate analysis (Table 2), only MIBC (hazard ratio [HR] = 21.1, 95% confidence interval [CI] = 4.24–105.1, p = 0.0001) proved to be an independent risk factor for DSS, resulting in a significant reduction in survival. Furthermore, MIBC was associated with a significant reduction in overall survival (HR = 9.60, CI: 2.59-35,5, p = 0.001).