RESUMO
PURPOSE: To assess the accuracy of prostate histoscanning (PHS) for spatial detection and localization of prostate cancer (PCa). PATIENTS AND METHODS: Prospective, single center study from January to September 2012 was conducted. Inclusion criterion was biopsy confirmed PCa in patients scheduled for radical prostatectomy. In total, 98 patients were included in the study. Results of PHS were compared against whole-mount step sectioning by the Stanford technique. A lower limit of 0.1 cm(3) was used for PHS. A dedicated 12-sector form was used for spatial correlation. The urologist and pathologist were blinded for each other's results. Sensitivity, specificity, and receiver operating characteristic curves were calculated with a logistic regression model for covariates. RESULTS: PHS performance for detection of PCa lesions ≥0.1 cm(3) had sensitivity of 60%, specificity of 66%, and area under the curve (AUC) of 0.63. Posterior and anterior sectors achieved sensitivity of 77%, specificity of 39%, and 28% and 84%, respectively. The model containing PHS positivity within a given sector reached sensitivity of 73.4%, specificity of 65.7%, and AUC of 0.75. In a ogistic regression model, the performance of PHS was affected by sector location, rectal distance, index, and total cancer volume (all P<0.0001) and bladder fullness (P=0.02). The best PHS accuracy was present in midposterior sectors. CONCLUSIONS: PHS has a potential for clinical practice, especially if PHS positivity within given sectors is taken into account. A trained operator is important. More studies are necessary to test different detection limits in various clinical settings, such as targeted biopsies and image guided focal therapy.