Introduction: The direct visualization of the biopsy needle and of the MRI-visible lesion provided by in-bore MRI-guided biopsies (IBBx) may result in superior sampling accuracy of these lesions compared to other techniques. This super-selective targeting, however, may oversample specific regions within a lesion with the potential for inadvertent risk inflation (e.g., overestimating the % of Gleason pattern 4 in a grade group [GG] 2 cancer). Conversely, the concurrent systematic sampling obtained during a MRI-TRUS fusion biopsy (FBx) may identify MRI-invisible lesions that would otherwise not be detected with IBBx. Preoperative GG revealed by each targeting approach, thus, may have different concordance rates with radical prostatectomy (RP). The goal of this study was to determine and compare the rates of GG concordance between preoperative targeted (IBBx vs FBx) biopsies and RP in men undergoing IBBx vs FBx.
Methods: This single-center, retrospective analysis of prospectively generated data included all men with abnormal mpMRI followed by IBBx or FBx between May/2017-Jan/2019 and May/2017-April/2018, respectively, who then underwent RP. FBx included targeted and systematic sampling. For assessment of GG agreement, the highest GG on the biopsy specimens was compared to the index lesion GG on the RP specimen at a patient-level. Two-sided two-proportion z-tests were used with significance level set at 0.05.
Results: 191 men (90 IBBx and 101 FBx; mean age 65y, PSA 8.3 ng/mL, prostate volume 53 cc, PSA density 0.18 ng/mL/cc) were eligible. Differences in concordance (IBBx: 67%, 60/90, FBx: 57%, 57/101) and downgrade (IBBx: 20%, 18/90; FBx: 16%, 16/101) rates were not statistically significant (p=0.16 and 0.47, respectively). There were fewer upgrades in the IBBx (13%,12/90) than in the FBx (28%, 28/101) group (p=0.01). Most (62%,46/74) reclassified cases involved GG 2-3 changes (Figure 1).
Conclusions: IBBx was associated with lower incidence of GG upgrade after RP compared to FBx. Concordance and downgrade rates were not significantly different between the two targeting approaches. Source of