Int J Urol 2017 02 8;24(2):137-143. Epub 2016 Nov 8.
Division of Urology, Department of Surgery, Duke University Medical Center, Durham, North Carolina, USA.
Objective: To evaluate the diagnostic properties of multiparametric magnetic resonance imaging in the detection, localization and characterization of prostate cancer using three-dimensional transperineal template mapping biopsy histopathology as the comparator.
Methods: A retrospective analysis of patients undergoing prostate multiparametric magnetic resonance imaging followed by three-dimensional transperineal template mapping biopsy was carried out. For imaging and pathology data, the prostate was divided in octants with the urethra being the midline. The index test properties were calculated using the biopsy histopathology as the reference test with the following end-points: any cancer, any Gleason ≥7, any Gleason ≥7 or cancer length of ≥4 mm and any Gleason ≥7 or 6 mm in any given core. The latter two definitions correspond to 0.2 and 0.5 mL of cancer volume, respectively. Diagnostic properties including sensitivity, specificity, positive and negative predictive values were calculated.
Results: A total of 50 patients were included in the study. A median of 55 (interquartile range 42-63) biopsy cores were obtained per patient. Of 400 prostate octants evaluated, 28.5% had prostate cancer on mapping biopsy, whereas 23% of octants were considered suspicious for cancer on imaging. Multiparametric magnetic resonance imaging negative predictive values for Gleason ≥7 and clinically significant cancers were 84-100%. Similarly, specificity ranged between 79% and 85%. Sensitivity and positive predictive value remained moderate for all the reference test definitions.
Conclusions: Multiparametric magnetic resonance imaging is a useful minimally-invasive tool for detection, localization and characterization of prostate cancer. This imaging modality has high negative predictive value and specificity, and therefore it could be used to reliably rule out clinically significant cancer, obviating the multicore mapping biopsy.