Mp05-20 Histopathologic Correlation of Pirads Version 2 Lesions on 3t Multiparametric Prostate Mri

Abstract

INTRODUCTION AND OBJECTIVES:The Prostate Imaging-Reporting and Data System (PIRADS) score was developed to evaluate the likelihood of malignancy for lesions seen in the peripheral zone and transition zone on multi-parametric Magnetic Resonance Imaging (MRI) of the prostate. We aim to determine if this same scoring system can be used to evaluate central zone lesions on MRI.METHODS: A retrospective review was performed of 73 patients who underwent MR/US fusion-guided biopsy of 139 suspicious lesions between February 2014 and October 2015. All patients underwent a 3 Tesla multi-parametric MRI. Indications for MRI include an abnormal digital rectal exam, PSA velocity >0.75ng/dl/year and patients on active surveillance. Our multi-parametric MRI sequence involved T2, diffusion weighted imaging and dynamic contrast enhancement. Using a 3-dimensional model software [InVivo (Phillips), Gainesville (USA)], at least 3 MR/US fusion-guided biopsies were performed on each prostate lesion seen on MRI regardless of PIRADS score under local anesthesia in the outpatient clinic.RESULTS: There were 80 peripheral zone lesions, 32 transitional zone lesions and 27 central zone lesions that were biopsied.Median PIRADS score for central zone lesions was 3 (range 1-5). Compared to the peripheral and transition zone, central zone lesions graded PIRADS 4 and 5 were more likely to be false positive, p¼0.012. Only two patients (7%) had clinically significant prostate cancer (Gleason >3+3) seen on central zone lesion. Both patients had lesions which were graded as PIRADS 3. Both lesions involved the transition zone as well and encompassed at least 50% of the entire central zone and transition zone. Both patients previously had transrectal ultrasound guided biopsy of the prostates which were negative for cancer. Both patients underwent a robotic assisted laparoscopic prostatectomy which yielded a Gleason score that were similar to MRI fusion biopsy.CONCLUSIONS:Lesions involving only the central zone seen on multi-parametric MRI are less concerning for malignancy and should not be given equal weightage as peripheral zone lesions. In our series, no lesions involving solely the central zone, regardless of PIRADS score was positive for malignancy on MR/US fusion-guided biopsy. A better PIRADS scoring system should be developed to help identify central zone lesions with malignant potential.