Between January 2008 and December 2012, a total of 128 PCa patients who underwent radical prostatectomy at Peking University Shougang Hospital and Peking University First Hospital (Beijing, China) were included in the present study. These patients did not receive radiotherapy or chemotherapy prior to surgery and clinical follow up data was available. The median time of postoperative follow-up was 35 months (range, 12–60 months). All the PCa cases were stratified into the following three groups according to the criteria in 2010 National Comprehensive Cancer Network clinical practice guidelines (18) on PCa: Low-risk, ≤cT2a, PSA <10 ng/ml, Gleason score ≤6; intermediate-risk, cT2b, PSA10-20 ng/ml, Gleason score=7; and high-risk, ≥cT2c, PSA >20 ng/ml, Gleason score ≥8. Clinicopathological data including patient age, pathological stage (tumor-node-metastasis classification system) and surgical margin status, was also recorded. Following tissue collection, the specimens were fixed in 10% neutral-buffered formalin (Bestbio Biotechnology Co., Ltd., Shanghai, China) and embedded in paraffin (Yijie Biotechnology Co., Ltd., Shanghai, China). Pathological diagnosis was performed preoperatively and confirmed postoperatively. The present study received approval by the Clinical Research Ethics Committee of the Peking University Shougang Hospital. Written informed consent was obtained from all of the patients.

The end point of follow-up study was the time of biochemical recurrence and was defined as a serum PSA level ≥0.2 ng/ml on two successive measurements performed 24 h apart (19). The period of biochemical recurrence-free survival was defined as the time interval from the date of surgery to the first day of biochemical recurrence.

All the tissue microarrays were constructed as previously described (20). In brief, one tissue core (0.6 mm) was taken from an individual paraffin-embedded PCa specimen (donor block) and arrayed precisely into a new paraffin block (35×20 mm; Zhongshan Golden Bridge Biotechnology Co. Ltd., Beijing, China) using a custom-built precision instrument (ATA-27; Beecher Instrument, Inc. Silver spring, MD, USA). Following block construction, 4 µm sections were cut from the microarray blocks using a Leica microtome (Leica RM2135; Leica Instruments GmbH, Hubloch, Germany) to support the adhesion of array elements. Overall, the tissue microarray block contained 128 donor blocks from specimens of all 128 PCa patients. The presence of prostate tissue on the arrayed specimens was verified via hematoxylin-eosin (Solarbio Biotechnology Co., Ltd., Shanghai, China) stained sectioning for the identification of pathological features associated with PCa tissue.

Tissue microarrays containing consecutive 4 µm sections were used for immunohistochemical staining. Staining was performed using theavidin-biotin complex (ABC) immunoperoxidase kit (SP-9000; Zhongshan Golden Bridge Biotechnology Co. Ltd.). Prior to antigen retrieval, deparaffinized sections were treated with methanol containing 3% hydrogen peroxide (Zhongshan Golden Bridge Biotechnology Co. Ltd.) for 10 min. Sections were then boiled for 5 min in a 0.01 M sodium citrate buffer (pH 6.0; Zhongshan Golden Bridge Biotechnology Co. Ltd.) in a water bath for antigen retrieval. Following washing with 0.01 M phosphate-buffered saline (PBS; ZLI-9062; Zhongshan Golden Bridge Biotechnology Co. Ltd.), sections were blocked with blocking serum (Zhongshan Golden Bridge Biotechnology Co. Ltd.) for 10 min at 37°C. The sections were incubated at 4°C overnight with rabbit monoclonal antibodies directed against human MT-2A (1:200; Sc11377), E-cadherin (1:150; Sc7870), cyclin E (1:100; Sc481), PCNA (1:200; Sc7907) and Bcl-2 (1:150; Sc783) as well as mouse monoclonal antibodies directed against human IL-6 (1:50; Sc1265), which were all purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). Additional negative controls lacking antibody were also used. The sections were then incubated with biotinylated goat anti-rabbit immunoglobulin G (SP-9000 Reagent B) for 20 min at 37°C and subsequently incubated with peroxidase-labeled streptavidin (SP-9000 Reagent C). Finally, 3,3′-diaminobenzidine (D12384; Sigma-Aldrich, St. Louis, MO, USA) was used as a chromogen and hematoxylin (ZLI-9609; Zhongshan Golden Bridge Biotechnology Co. Ltd.) was used as a counterstain. In order to determine the expression of these biomarkers in PCa cells, five microscopic fields were visualized in each core (Olympus BX40 microscope; Olympus Corporation, Tokyo, Japan). The percentage of cells with moderate staining was required in ≥10% to be considered as positive expression.

Statistical analyses were performed using the SPSS 13.0 software (SPSS, Inc., Chicago, IL, USA). Continuous variables were expressed as the mean ± standard deviations. Statistical analysis was performed independently using the Student's t-test, Fisher's exact test for any 2×2 tables and Pearson's χ² test for non-2×2 tables. The Cox proportional hazards model was used to evaluate the association of conventional clinicopathological parameters and biomarkers with biochemical recurrence in univariate and multivariate analysis. Univariate biochemical recurrence-free survival was tested using the Kaplan-Meier analysis and with log-rank test for difference. P<0.05 was considered to indicate a statistical significant difference between values.

Of 128 PCa cases, there were 54 low-risk, 62 intermediate-risk and 12 high-risk cases. The mean age of PCa patients was 67.2±10.3 years (range, 53–78 years), the mean preoperative serum PSA level was 13.8±8.9 ng/ml (range, 1.3–32.1 ng/ml), the mean Gleason score was 6.8±1.9 (range, 4–10) and the positive surgical margin rate was 15.6%.

In this study, the median biochemical recurrence-free time was 19 months (range, 6–35 months). The biochemical recurrence rate was 30.5% in all cases (39/128) and for the low-, intermediate- and high-risk classifications the biochemical recurrence rates were 14.8 (8/54), 38.7 (24/62) and 58.3% (7/12), respectively.

The results of immunostaining for MT-2A, E-cadherin, IL-6, cyclin E, PCNA and Bcl-2 in PCa tissues are shown in Fig. 1. Expression rates of MT-2A, E-cadherin, IL-6, cyclin E, PCNA and Bcl-2 were 53.9 (69/128), 46.9 (60/128), 46.1 (59/128), 50.8 (65/128), 49.2 (63/128) and 53.1% (68/128) in PCa cases, respectively. Furthermore, the risk classification and pathological T (pT) stage of PCa were found to be positively associated with MT-2A (P=0.042 and 0.015), IL-6 (P=0.031 and 0.019) and cyclin-E expression (P=0.039 and 0.043), but inversely correlated with E-cadherin expression (P=0.032 and 0.012) (Table I). In addition, pT stage was positively correlated with PCNA and Bcl-2 expression (P=0.026 and 0.011, respectively).

According to the univariate and multivariate Cox regression analysis, PCa risk classification was demonstrated to be an independent risk factor for biochemical recurrence [hazard ratio (HR)=1.81; 95% confidence interval (CI)=1.56–2.19; P=0.047] (Table II).

Based on the results of univariate analysis, the positive expression of MT-2A (HR=2.15; 95% CI=1.14–3.08; P=0.009) and cyclin E (HR=1.45; 95% CI=1.02–1.92; P=0.037) as well as the negative expression of E-cadherin (HR=1.31; 95% CI=1.03–2.21; P=0.047) were indicated to be significant prognostic factors for the biochemical recurrence of PCa (Table II). Furthermore, univariate Kaplan-Meier/log-rank analysis revealed that decreased biochemical recurrence-free survival was noted in PCa cases with positive expression of MT-2A and cyclin E as well as negative expression of E-cadherin (P=0.022, 0.028 and 0.011 log rank test, respectively) (Fig. 2).

According to multivariate Cox regression analysis, MT-2A expression (HR=2.01; 95% CI=1.08–3.15; P=0.015), E-cadherin expression (HR=1.79; 95% CI=1.08–2.21; P=0.042) and cyclin E expression (HR=1.92; 95% CI=1.22–2.45; P=0.020) were demonstrated to be independent predictors for biochemical recurrence of PCa (Table II).