Background. Study compares results of preoperative diagnostic methods with measurements of tumor extension in a specimen obtained by radical prostatectomy. Methods and Results. One hundred forty-six patients with clinically identified cancer were enrolled in the study and subjected to radical prostatectomy. Each patient underwent transrectal sonography (TRUS), Power Doppler sonography with 3D reconstruction (3D-PDS), and the risk of the occurrence of locally advanced tumor was assessed using Partin tables. In subgroups of localized and locally advanced tumor, individual preoperative parameters were compared. ROC curves were generated for individual preoperative parameters and the area under the curve (AUC) was calculated. Multivariate logistic regression analysis was performed to determine independent predictors of extraprostatic tumor extension. A statistically significant difference between patients with localized and locally advanced tumor was observed in PSA levels (p<0.014), PSA density (p<0.004), DRE (p<0.037), TRUS (p<0.003), and 3D-PDS (p<0.000). The highest AUC value was found for 3D-PDS 0.776, TRUS 0.670, and PSA density 0.639. According to multivariate analysis, independent preoperative predictors of extraprostatic tumor extension were PSA density, preoperative Gleason score 7, and 3D-PDS finding. Conclusions. Power Doppler sonography with 3D reconstruction represents the most reliable preoperative diagnostic tool in determining locally advanced tumor. Together with PSA density and the presence of aggressive tumor (Gleason score ≥7) in prostate biopsy, it is one of independent predictors of locally advanced prostate cancer.

        Key words: prostate cancer, diagnostics, radical prostatectomy