We evaluated the diagnostic accuracy of multiparametric magnetic resonance imaging to diagnose clinically significant prostate cancer and compared it with the diagnostic accuracy of transperineal saturation prostate biopsy.Materials and Methods:
From January 2011 to February 2018 repeat saturation prostate biopsy (the reference test) was done due to suspicion of cancer in 1,032 men with a median age of 63 years in whom median prostate specific antigen was 8.6 ng/ml. All patients underwent 3.0 Tesla pelvic multiparametric magnetic resonance imaging before saturation prostate biopsy. Additional targeted fusion prostate biopsy was done of lesions with a PI-RADS™ (Prostate Imaging Reporting and Data System) score of 3 or greater.Results:
T1c prostate cancer was found in 372 of the 1,032 patients (36%). Of these cases 272 (73.1%) were classified as clinically significant prostate cancer. Saturation prostate biopsy vs targeted fusion prostate biopsy and a PI-RADS score of 3 or greater vs targeted fusion prostate biopsy and a PI-RADS score of 4 or greater diagnosed 95.6% vs 83.8% vs 60.3% of clinically significant prostate cancers (p <0.0001). Saturation prostate biopsy missed 12 of 272 clinically significant prostate cancers (4.5%) vs 44 (16.2%) and 108 of 272 (39.7%) missed by targeted fusion prostate biopsy and a PI-RADS score of 3 or greater and a score of 4 or greater, respectively (p <0.0001). As a triage test multiparametric magnetic resonance imaging would have spared 49.3% vs 73.6% of patients using a PI-RADS cutoff of 3 or greater vs 4 or greater.Conclusions:
Multiparametric magnetic resonance imaging could significantly reduce the number of unnecessary repeat prostate biopsies in about 50% of cases in which a PI-RADS score of 3 or greater is used. At the same time patients should be informed of the 16.2% and 39.7% false-negative rates of clinically significant prostate cancer for targeted fusion prostate biopsy of PI-RADS 3 or greater and 4 or greater lesions, respectively.