Pathology results of 308 consecutive prostatectomies (2014-2018) were retrospectively evaluated. Prostate tumor variants and atypical findings in non-cancerous prostate tissue were selected for analysis. 

Pathological data was obtained from the prostatectomy reports elaborated by a specialized urological pathologist. Medical records were evaluated for patient age, lower urinary tract obstructive symptoms, prostate specific antigen (PSA) serum levels (ng/ml), signs of dissemination and disease history. The pre-surgical biopsy reports were also compared with the prostatectomy pathology results. 

MRI examinations were performed using a 1.5 Tesla (T) or a 3.0 T magnet (Siemens Medical Solutions, Erlangen, Germany). The MRI protocol consisted of an axial T1 and multiplanar T2-WI, DWI of the pelvis (b-values of 0, 50, 200, 500, 800 and 1000 s/mm2). A dynamic contrast-enhanced (DCE) sequence of the pelvis and an axial T2-WI of the retroperitoneum were also adquired. 

The MRI studies were reviewed by an urological radiologist with 10 years of experience and two training radiologists. The location of the pathology findings were correlated with the MRI images.

The following MRI parameters were evaluated: prostate volume; lesion´s location and largest diameter (mm); T1, T2, DWI and ADC intensity (qualitatively); lesion enhancement. Data collection was obtained by consensus. 

Patient’s clinical outcome was followed using their electronic medical records (mean time: 31 months). Biochemical and clinical recurrence were recorded as endpoints.