Not All Multiparametric Magnetic Resonance Imaging-targeted Biopsies Are Equal: The Impact of the Type of Approach and Operator Expertise on the Detection of Clinically Significant Prostate Cancer

Background: The extensive use of multiparametric magnetic resonance imaging (mpMRI) has led to an even more widespread use of different targeted biopsy techniques and approaches. The best way of performing targeted biopsies and the effect of operator expertise have still to be defined.

Objective: To compare the rate of detection of clinically significant prostate cancer (csPCa) of different mpMRI targeted approaches and to assess the role of operator expertise in the detection of csPCa.

Design, setting, and participants: We included 244 consecutive patients who underwent both 12-core transrectal ultrasound (TRUS) biopsy and mpMRI targeted biopsy with either a cognitive biopsy (CB) or fusion biopsy (FB) approach during the same session between 2013 and 2016 at a single tertiary referral centre.

Intervention: All men underwent 1.5-T mpMRI with an endorectal coil. All biopsies were performed by three operators as their first cases of targeted biopsy. Lesions with a Prostate Imaging Recording and Data System (PI-RADS) v.2 score of ≥3 detected at mpMRI were targeted. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: csPCa was defined as disease with a Gleason score at biopsy of ≥7. Operator expertise was coded as the progressive number of targeted biopsies performed by each physician. Multivariable logistic regression analyses (MVA) were used to assess the association between the targeted biopsy technique (FB vs CB) and operator expertise for detection of csPCa. Covariates consisted of prostate-specific antigen, prostate volume, PI-RADS v.2 (3 vs >3), number of targeted cores per MRI lesion, and digital rectal examination (negative vs positive). The same analyses were performed for patients undergoing FB only after accounting for the FB approach (transrectal vs transperineal). A lowess smoothing weighted function was used to graphically assess the effect of operator expertise on the probability of detecting csPCa, after accounting for all confounders.

Results and limitations: Overall, 157 patients (64%) underwent FB and 87 (36%) underwent CB. The overall csPCa detection rate was 58% for FB and 45% for CB (p=0.07). A significantly higher rate of csPCa detection in targeted samples was observed for FB compared to CB (57% vs 36%; p=0.002). On MVA, FB and operator expertise were significantly associated with a higher probability of csPCa detection in targeted samples (odds ratio [OR] 2.4 and 1.7, respectively; both p≤0.03). When the same analyses were repeated for patients undergoing FB, operator expertise remained an independent predictor of csPCa detection (OR 1.9; p=0.004). An increase in the probability of detecting csPCa with the number of procedures performed was observed after accounting for all confounders.

Conclusions: We demonstrated that FB had higher detection rate than CB for csPCa. Moreover, operator expertise was significantly associated with higher detection rates for csPCa.

Patient summary: When different targeted biopsy techniques were compared, fusion biopsy provided a higher detection rate compared to cognitive biopsy for clinically significant prostate cancer (csPCa). Moreover, we found that operator expertise was an important predictor of the detection of csPCa, regardless of the procedure used.