A total of 268 patients aged 71±8 years old and suspected of having PCa were admitted and treated at Ruijin Hospital (Shanghai, China) between January 2013 and December 2013. Indications for prostate biopsy were as follows: Serum PSA >4 ng/ml, or ultrasound or MRI found suspicious nodules, or clinical anus examination identified hard nodules. All patients were hospitalized, underwent routine hospitalization tests and coagulation index detection. A total of 6 prostate biopsy specimens were obtained from each patient and all specimens possessed pathological results. Prior written and informed consent was obtained from each participant and the present study was approved by the Ethics Committee of Ruijin Hospital Affiliated to Shanghai Jiaotong University, School of Medicine (Shanghai, China).

Patients should have not ejaculated within 24 h, conducted bladder microscopy or catheterization within 48 h, accepted prostate massage within 1 week or performed prostate puncture within 1 month prior to serum PSA evaluation. This was to limit the presence of confounding factors that may have affected PSA levels. Following overnight fasting, patients had 3 ml forearm venous blood drawn in the morning and were inspected using a PSA ELISA kit (cat. no. KA0208; Abnova, Taipei, Taiwan R.O.C.) and analyzed using an automated immunochemistry analyzer (Abbott Laboratories, Abbott Park, IL, USA).

A MyLab^™ 60 diasonograph (Esaote, Genova, Italy), rectal biplanar probe (Esaote), automated biopsy gun (Bard Biopsy Systems, Tempe, AZ, USA) and 16-gauge (G) Tru-cut biopsy needle (Bard Biopsy Systems) were used in the present study. The length of the specimen slot was set to 2.2 cm. Patients were asked to lift up the scrotum to fully expose the perineum, then transverse scanning was performed using a convex array probe to measure the prostate anteroposterior and transverse diameters. Axial scanning was performed by linear array probe to measure prostate axial diameter. This allowed for the detection of nodules, cysts or calcification, and the selection of an appropriate puncture site.

Following perineum disinfection, infiltration anesthesia was administered with 5 ml 1% lidocaine (Century Pharmaceutical Co., Ltd., Tianjin, China) on each side of the midcourt line, which was at the middle of the urethral bulb and ischial tuberosity, and 1.5–2.0 cm away from the anus. The sextant biopsy scheme was subsequently utilized to obtain prostate biopsy specimens. Puncture points were on the tip of the prostate near the midline sagittal section, as well as the middle and bottom of the prostrate, for a total of six puncture points per patient.

One sonographer, one urologist and one physician pathologist evaluated all the specimens. Specimens were scored based on five aspects: Hardness, color, plumpness, transparency and uniformity, and were given a score of either 0 or 1 (Table I). Specimen scores were recorded under double-blinded conditions and thereafter, the averages of six specimen ratings were calculated.

Prostate tissues were fixed with 4% paraformalde-hyde at room temperature for 24 h, embedded in paraffin and then sliced into sections (6 µm). Briefly, slices were dewaxed with xylene and rehydrated using graded alcohols. Following rinsing with tap water and distilled water, slices were stained with hematoxylin for 3–5 min. Slices were washed with tap water and stained with eosin for 1–4 min and subsequently underwent dehydration and differentiation with HCl in 70% alcohol. Slices were then studied using a light microscope.

Quantitative data were expressed as the mean ± standard error of the mean. Statistical comparisons between the PCa and non-PCa groups were performed using a Student's t-test. Logistic regression and receiver operator characteristic curve (ROC) analysis were used to analyze the sensitivity and specificity of the visually determined score and PSA level. Finally, Spearman correlation was applied to determine the correlation between the visually determined scores and PSA levels. P<0.05 was considered to indicate a statistically significant difference.

The average age of the 268 patients with suspected PCa patients was 71±8 years old. The indications for prostate biopsy were serum PSA >4 ng/ml, a suspicious nodule identified by heterogeneous echo that was confirmed by ultrasound/magnetic resonance imaging examination or hard nodules revealed following clinical rectal examination. The demographic and clinical characteristics of all patients are presented in Table II.

The corresponding H&E staining of the PCa puncture specimen is presented in Fig. 1. The following features were observed: Irregular glands, glandular fibrous hyperplasia, tumor invasive growth, basal cell disappearance, tumor cell nucleolus appearance, mostly single cell morphology.

A representative extirpated tumor from a PCa patient is presented in Fig. 2. The tumor can be seen with the naked eye. The tumor had many ashen hard nodules from 0.3–1.3 cm in diameter with some red patches spreading out.

PSA levels in the PCa group (99.35±1.40) were significantly higher than those of the control group (17.12±0.10; Fig. 3A; P<0.01). However, the visually determined score (3.44±0.01) of the PCa group was also significantly increased compared with the control group (0.99±0.01; Fig. 3B; P<0.01). Correlation analysis was carried out on the biopsy specimen visual grading score and PSA. The correlation coefficient was 0.255 (P<0.001), which indicated a linear relationship between the two characteristics.

Logistic regression analysis verified that upregulated PSA and visually determined scores were two independent determining factors in the diagnosis of PCa (Table III). From the data it was identified that PAS and prostate biopsy specimen visual grading have statistical significance to distinguish between prostate cancer and non-prostate cancer patients. Higher PSA (OR=1.045, 95% CI=1.021–1.069, P<0.001) or higher visual grading score (OR=14.096, 95% CI=6.326–31.409, P<0.001) was correlated with a higher risk of PCa.

An ROC curve analysis was performed to further investigate the clinical value of PSA and visually determined score in the diagnosis of PCa. As shown in Fig. 4 and Table IV, the area under ROC curve of visual grading biopsy specimens was 0.929 (P<0.001). The area under ROC curve of PSA was 0.817 (P<0.001). The area under ROC curve for visual grading was significantly higher compared with PSA (P<0.05). Overall, this indicates that the visually determined score is more accurate than PSA in the clinical diagnosis of PCa.

In the current study, if visually determined score was greater than 1, 2 or 3 scores, it was considered to be positive. The sensitivity and specificity of the visually determined score in the diagnosis of PCa is presented in Table V. Taking sensitivity and specificity into consideration, diagnosis of PCa using the visual method should take into account 2 scores (the sensitivity is 0.817 and the specificity is 0.931), as this has the highest sensitivity and specificity compared with other numbers of scores.

According to the visual method of two scores, patients were divided into two groups as shown in Table VI. The table showed that when the cutoff for diagnosis used as the visually determined score was set to ≥2 points, the positive predictive value of the diagnosis was 61.8% whereas the negative predictive value of the diagnosis was 96.8%.