ABSTRACT
BACKGROUND: The objective of this study was to analyze the potential of prostate magnetic resonance imaging (MRI) and MRI/transrectal ultrasound-fusion biopsies to detect and to characterize significant prostate cancer (sPC) in the anterior fibromuscular stroma (AFMS) and in the transition zone (TZ) of the prostate and to assess the accuracy of multiparametric MRI (mpMRI) and biparametric MRI (bpMRI) (T2w and diffusion-weighted imaging (DWI)). METHODS: Seven hundred and fifty-five consecutive patients underwent prebiopsy 3 T mpMRI and transperineal biopsy between October 2012 and September 2014. MRI images were analyzed using PIRADS (Prostate Imaging-Reporting and Data System). All patients had systematic biopsies (SBs, median n=24) as reference test and targeted biopsies (TBs) with rigid software registration in case of MRI-suspicious lesions. Detection rates of SBs and TBs were assessed for all PC and sPC patients defined by Gleason score (GS)⩾3+4 and GS⩾4+3. For PC, which were not concordantly detected by TBs and SBs, prostatectomy specimens were assessed. We further compared bpMRI with mpMRI. RESULTS: One hundred and ninety-one patients harbored 194 lesions in AFMS and TZ on mpMRI. Patient-based analysis detected no difference in the detection of all PC for SBs vs TBs in the overall cohort, but in the repeat-biopsy population TBs performed significantly better compared with SBs (P=0.004 for GS⩾3+4 and P=0.022 for GS⩾4+3, respectively). Nine GS⩾4+3 sPCs were overlooked by SBs, whereas TBs missed two sPC in men undergoing primary biopsy. The combination of SBs and TBs provided optimal local staging. Non-inferiority analysis showed no relevant difference of bpMRI to mpMRI in sPC detection. CONCLUSIONS: MRI-targeted biopsies detected significantly more anteriorly located sPC compared with SBs in the repeat-biopsy setting. The more cost-efficient bpMRI was statistically not inferior to mpMRI in sPC detection in TZ/AFMS.
