3T MRI of the prostate correlation to 3D reconstruction of transperineal mapping biopsies.
e15176 Background: MRI is new tool in the diagnosis and surveillance of localized prostate cancer. Studies have examined the sensitivity and specificity of MRI for detecting and localizing prostate cancer as compared to radical prostatectomy specimens. We have previously demonstrated excellent correlation between 3-dimensional transperineal mapping biopsies (3DMB) to radical prostatectomy specimens. We hypothesize that MRI does not perform as well detecting smaller lesions of the prostate as larger ones. In this study, we retrospectively compare the correlation of 3DMB to 3T MRI-positive lesions. Methods: Between 2010 and 2011, 34 men with histologically-proven prostate cancer underwent 3T endorectal coil MRI and subsequent template-guided (5-mm grid) 3DMB. Three-dimensional computer models of the mapping biopsies were created using Proview software to identify clusters of positive biopsies likely belonging to a single lesion. MRIs were read by two trained radiologists blinded to 3DMB results. Disagreement between radiologists was settled by consensus. MRI and 3DMB were then compared in a retrospective fashion. Results: Median age and PSA were 65 years (49-76) and 4.3 ng/mL (0.7-15.0). Median prostate volume was 40 cm3 (21-105 cm3). Median total biopsy cores on 3DMB was 60 (26-151), and mean positive cores was 3 (0-13). 3D reconstruction of patient biopsy sets resulted in 58 distinct foci of cancer. MRI identified 64 individual lesions suspicious for cancer. Of these, 66% (n=42) could be correlated to a positive area on 3DMB consisting of one or more positive biopsies, and 34% (n=22) corresponded to negative regions on 3DMB. When two or more biopsies represented a positive lesion on 3DMB, the positive-predictive value of MRI was 100% (n=24). Conversely, only 72% (n=42) of positive 3DMB areas could be correlated to a lesion on MRI, and 28% (n=16) were not seen on MRI.. Conclusions: MRI has vast potential as a less-invasive method in the detection and surveillance of prostate cancer. While MRI does appear to have excellent positive predictive value, the number of false negatives and relevance to clinically-significant lesions require further illumination.